diff options
154 files changed, 2751 insertions, 1545 deletions
diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index c2992bc54f2..8b22d7d8b99 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -92,7 +92,6 @@ prototypes: void (*destroy_inode)(struct inode *); void (*dirty_inode) (struct inode *); int (*write_inode) (struct inode *, int); - void (*put_inode) (struct inode *); void (*drop_inode) (struct inode *); void (*delete_inode) (struct inode *); void (*put_super) (struct super_block *); @@ -115,7 +114,6 @@ alloc_inode: no no no destroy_inode: no dirty_inode: no (must not sleep) write_inode: no -put_inode: no drop_inode: no !!!inode_lock!!! delete_inode: no put_super: yes yes no diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index 81e5be6e6e3..b7522c6cbae 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -205,7 +205,6 @@ struct super_operations { void (*dirty_inode) (struct inode *); int (*write_inode) (struct inode *, int); - void (*put_inode) (struct inode *); void (*drop_inode) (struct inode *); void (*delete_inode) (struct inode *); void (*put_super) (struct super_block *); @@ -246,9 +245,6 @@ or bottom half). inode to disc. The second parameter indicates whether the write should be synchronous or not, not all filesystems check this flag. - put_inode: called when the VFS inode is removed from the inode - cache. - drop_inode: called when the last access to the inode is dropped, with the inode_lock spinlock held. diff --git a/Documentation/s390/CommonIO b/Documentation/s390/CommonIO index 8fbc0a85287..bf0baa19ec2 100644 --- a/Documentation/s390/CommonIO +++ b/Documentation/s390/CommonIO @@ -8,17 +8,6 @@ Command line parameters Enable logging of debug information in case of ccw device timeouts. - -* cio_msg = yes | no - - Determines whether information on found devices and sensed device - characteristics should be shown during startup or when new devices are - found, i. e. messages of the types "Detected device 0.0.4711 on subchannel - 0.0.0042" and "SenseID: Device 0.0.4711 reports: ...". - - Default is off. - - * cio_ignore = {all} | {<device> | <range of devices>} | {!<device> | !<range of devices>} diff --git a/MAINTAINERS b/MAINTAINERS index abe27871ad6..f5583dc7ea3 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2112,12 +2112,10 @@ L: netdev@vger.kernel.org S: Maintained INTEL ETHERNET DRIVERS (e100/e1000/e1000e/igb/ixgb/ixgbe) -P: Auke Kok -M: auke-jan.h.kok@intel.com -P: Jesse Brandeburg -M: jesse.brandeburg@intel.com P: Jeff Kirsher M: jeffrey.t.kirsher@intel.com +P: Jesse Brandeburg +M: jesse.brandeburg@intel.com P: Bruce Allan M: bruce.w.allan@intel.com P: John Ronciak diff --git a/arch/cris/kernel/sys_cris.c b/arch/cris/kernel/sys_cris.c index 8b9984197ed..d124066e172 100644 --- a/arch/cris/kernel/sys_cris.c +++ b/arch/cris/kernel/sys_cris.c @@ -40,8 +40,11 @@ asmlinkage int sys_pipe(unsigned long __user * fildes) error = do_pipe(fd); unlock_kernel(); if (!error) { - if (copy_to_user(fildes, fd, 2*sizeof(int))) + if (copy_to_user(fildes, fd, 2*sizeof(int))) { + sys_close(fd[0]); + sys_close(fd[1]); error = -EFAULT; + } } return error; } diff --git a/arch/m32r/kernel/sys_m32r.c b/arch/m32r/kernel/sys_m32r.c index 6d7a80fdad4..319c79720b8 100644 --- a/arch/m32r/kernel/sys_m32r.c +++ b/arch/m32r/kernel/sys_m32r.c @@ -90,8 +90,11 @@ sys_pipe(unsigned long r0, unsigned long r1, unsigned long r2, error = do_pipe(fd); if (!error) { - if (copy_to_user((void __user *)r0, fd, 2*sizeof(int))) + if (copy_to_user((void __user *)r0, fd, 2*sizeof(int))) { + sys_close(fd[0]); + sys_close(fd[1]); error = -EFAULT; + } } return error; } diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 29a7940f284..1d035082e78 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -430,6 +430,13 @@ config CMM_IUCV Select this option to enable the special message interface to the cooperative memory management. +config PAGE_STATES + bool "Unused page notification" + help + This enables the notification of unused pages to the + hypervisor. The ESSA instruction is used to do the states + changes between a page that has content and the unused state. + config VIRT_TIMER bool "Virtual CPU timer support" help diff --git a/arch/s390/kernel/compat_wrapper.S b/arch/s390/kernel/compat_wrapper.S index 743d54f0b8d..d003a6e16af 100644 --- a/arch/s390/kernel/compat_wrapper.S +++ b/arch/s390/kernel/compat_wrapper.S @@ -121,7 +121,7 @@ sys32_ptrace_wrapper: lgfr %r3,%r3 # long llgtr %r4,%r4 # long llgfr %r5,%r5 # long - jg sys_ptrace # branch to system call + jg compat_sys_ptrace # branch to system call .globl sys32_alarm_wrapper sys32_alarm_wrapper: diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index bdbb3bcd78a..708cf9cf9a3 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -279,8 +279,6 @@ sysc_do_restart: st %r2,SP_R2(%r15) # store return value (change R2 on stack) sysc_return: - tm SP_PSW+1(%r15),0x01 # returning to user ? - bno BASED(sysc_restore) tm __TI_flags+3(%r9),_TIF_WORK_SVC bnz BASED(sysc_work) # there is work to do (signals etc.) sysc_restore: @@ -312,6 +310,8 @@ sysc_work_loop: # One of the work bits is on. Find out which one. # sysc_work: + tm SP_PSW+1(%r15),0x01 # returning to user ? + bno BASED(sysc_restore) tm __TI_flags+3(%r9),_TIF_MCCK_PENDING bo BASED(sysc_mcck_pending) tm __TI_flags+3(%r9),_TIF_NEED_RESCHED @@ -602,12 +602,6 @@ io_no_vtime: la %r2,SP_PTREGS(%r15) # address of register-save area basr %r14,%r1 # branch to standard irq handler io_return: - tm SP_PSW+1(%r15),0x01 # returning to user ? -#ifdef CONFIG_PREEMPT - bno BASED(io_preempt) # no -> check for preemptive scheduling -#else - bno BASED(io_restore) # no-> skip resched & signal -#endif tm __TI_flags+3(%r9),_TIF_WORK_INT bnz BASED(io_work) # there is work to do (signals etc.) io_restore: @@ -629,10 +623,18 @@ io_restore_trace_psw: .long 0, io_restore_trace + 0x80000000 #endif -#ifdef CONFIG_PREEMPT -io_preempt: +# +# switch to kernel stack, then check the TIF bits +# +io_work: + tm SP_PSW+1(%r15),0x01 # returning to user ? +#ifndef CONFIG_PREEMPT + bno BASED(io_restore) # no-> skip resched & signal +#else + bnz BASED(io_work_user) # no -> check for preemptive scheduling + # check for preemptive scheduling icm %r0,15,__TI_precount(%r9) - bnz BASED(io_restore) + bnz BASED(io_restore) # preemption disabled l %r1,SP_R15(%r15) s %r1,BASED(.Lc_spsize) mvc SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15) @@ -646,10 +648,7 @@ io_resume_loop: br %r1 # call schedule #endif -# -# switch to kernel stack, then check the TIF bits -# -io_work: +io_work_user: l %r1,__LC_KERNEL_STACK s %r1,BASED(.Lc_spsize) mvc SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15) diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index 5a4a7bcd2bb..fee10177dbf 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -271,8 +271,6 @@ sysc_noemu: stg %r2,SP_R2(%r15) # store return value (change R2 on stack) sysc_return: - tm SP_PSW+1(%r15),0x01 # returning to user ? - jno sysc_restore tm __TI_flags+7(%r9),_TIF_WORK_SVC jnz sysc_work # there is work to do (signals etc.) sysc_restore: @@ -304,6 +302,8 @@ sysc_work_loop: # One of the work bits is on. Find out which one. # sysc_work: + tm SP_PSW+1(%r15),0x01 # returning to user ? + jno sysc_restore tm __TI_flags+7(%r9),_TIF_MCCK_PENDING jo sysc_mcck_pending tm __TI_flags+7(%r9),_TIF_NEED_RESCHED @@ -585,12 +585,6 @@ io_no_vtime: la %r2,SP_PTREGS(%r15) # address of register-save area brasl %r14,do_IRQ # call standard irq handler io_return: - tm SP_PSW+1(%r15),0x01 # returning to user ? -#ifdef CONFIG_PREEMPT - jno io_preempt # no -> check for preemptive scheduling -#else - jno io_restore # no-> skip resched & signal -#endif tm __TI_flags+7(%r9),_TIF_WORK_INT jnz io_work # there is work to do (signals etc.) io_restore: @@ -612,10 +606,41 @@ io_restore_trace_psw: .quad 0, io_restore_trace #endif -#ifdef CONFIG_PREEMPT -io_preempt: +# +# There is work todo, we need to check if we return to userspace, then +# check, if we are in SIE, if yes leave it +# +io_work: + tm SP_PSW+1(%r15),0x01 # returning to user ? +#ifndef CONFIG_PREEMPT +#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE) + jnz io_work_user # yes -> no need to check for SIE + la %r1, BASED(sie_opcode) # we return to kernel here + lg %r2, SP_PSW+8(%r15) + clc 0(2,%r1), 0(%r2) # is current instruction = SIE? + jne io_restore # no-> return to kernel + lg %r1, SP_PSW+8(%r15) # yes-> add 4 bytes to leave SIE + aghi %r1, 4 + stg %r1, SP_PSW+8(%r15) + j io_restore # return to kernel +#else + jno io_restore # no-> skip resched & signal +#endif +#else + jnz io_work_user # yes -> do resched & signal +#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE) + la %r1, BASED(sie_opcode) + lg %r2, SP_PSW+8(%r15) + clc 0(2,%r1), 0(%r2) # is current instruction = SIE? + jne 0f # no -> leave PSW alone + lg %r1, SP_PSW+8(%r15) # yes-> add 4 bytes to leave SIE + aghi %r1, 4 + stg %r1, SP_PSW+8(%r15) +0: +#endif + # check for preemptive scheduling icm %r0,15,__TI_precount(%r9) - jnz io_restore + jnz io_restore # preemption is disabled # switch to kernel stack lg %r1,SP_R15(%r15) aghi %r1,-SP_SIZE @@ -629,10 +654,7 @@ io_resume_loop: jg preempt_schedule_irq #endif -# -# switch to kernel stack, then check TIF bits -# -io_work: +io_work_user: lg %r1,__LC_KERNEL_STACK aghi %r1,-SP_SIZE mvc SP_PTREGS(__PT_SIZE,%r1),SP_PTREGS(%r15) @@ -653,6 +675,11 @@ io_work_loop: j io_restore io_work_done: +#if defined(CONFIG_KVM) || defined(CONFIG_KVM_MODULE) +sie_opcode: + .long 0xb2140000 +#endif + # # _TIF_MCCK_PENDING is set, call handler # diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 7f427016374..35827b9bd4d 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -292,8 +292,7 @@ poke_user(struct task_struct *child, addr_t addr, addr_t data) return 0; } -static int -do_ptrace_normal(struct task_struct *child, long request, long addr, long data) +long arch_ptrace(struct task_struct *child, long request, long addr, long data) { ptrace_area parea; int copied, ret; @@ -529,35 +528,19 @@ poke_user_emu31(struct task_struct *child, addr_t addr, addr_t data) return 0; } -static int -do_ptrace_emu31(struct task_struct *child, long request, long addr, long data) +long compat_arch_ptrace(struct task_struct *child, compat_long_t request, + compat_ulong_t caddr, compat_ulong_t cdata) { - unsigned int tmp; /* 4 bytes !! */ + unsigned long addr = caddr; + unsigned long data = cdata; ptrace_area_emu31 parea; int copied, ret; switch (request) { - case PTRACE_PEEKTEXT: - case PTRACE_PEEKDATA: - /* read word at location addr. */ - copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 0); - if (copied != sizeof(tmp)) - return -EIO; - return put_user(tmp, (unsigned int __force __user *) data); - case PTRACE_PEEKUSR: /* read the word at location addr in the USER area. */ return peek_user_emu31(child, addr, data); - case PTRACE_POKETEXT: - case PTRACE_POKEDATA: - /* write the word at location addr. */ - tmp = data; - copied = access_process_vm(child, addr, &tmp, sizeof(tmp), 1); - if (copied != sizeof(tmp)) - return -EIO; - return 0; - case PTRACE_POKEUSR: /* write the word at location addr in the USER area */ return poke_user_emu31(child, addr, data); @@ -587,82 +570,11 @@ do_ptrace_emu31(struct task_struct *child, long request, long addr, long data) copied += sizeof(unsigned int); } return 0; - case PTRACE_GETEVENTMSG: - return put_user((__u32) child->ptrace_message, - (unsigned int __force __user *) data); - case PTRACE_GETSIGINFO: - if (child->last_siginfo == NULL) - return -EINVAL; - return copy_siginfo_to_user32((compat_siginfo_t - __force __user *) data, - child->last_siginfo); - case PTRACE_SETSIGINFO: - if (child->last_siginfo == NULL) - return -EINVAL; - return copy_siginfo_from_user32(child->last_siginfo, - (compat_siginfo_t - __force __user *) data); } - return ptrace_request(child, request, addr, data); + return compat_ptrace_request(child, request, addr, data); } #endif -long arch_ptrace(struct task_struct *child, long request, long addr, long data) -{ - switch (request) { - case PTRACE_SYSCALL: - /* continue and stop at next (return from) syscall */ - case PTRACE_CONT: - /* restart after signal. */ - if (!valid_signal(data)) - return -EIO; - if (request == PTRACE_SYSCALL) - set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); - else - clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); - child->exit_code = data; - /* make sure the single step bit is not set. */ - user_disable_single_step(child); - wake_up_process(child); - return 0; - - case PTRACE_KILL: - /* - * make the child exit. Best I can do is send it a sigkill. - * perhaps it should be put in the status that it wants to - * exit. - */ - if (child->exit_state == EXIT_ZOMBIE) /* already dead */ - return 0; - child->exit_code = SIGKILL; - /* make sure the single step bit is not set. */ - user_disable_single_step(child); - wake_up_process(child); - return 0; - - case PTRACE_SINGLESTEP: - /* set the trap flag. */ - if (!valid_signal(data)) - return -EIO; - clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); - child->exit_code = data; - user_enable_single_step(child); - /* give it a chance to run. */ - wake_up_process(child); - return 0; - - /* Do requests that differ for 31/64 bit */ - default: -#ifdef CONFIG_COMPAT - if (test_thread_flag(TIF_31BIT)) - return do_ptrace_emu31(child, request, addr, data); -#endif - return do_ptrace_normal(child, request, addr, data); - } - /* Not reached. */ - return -EIO; -} - asmlinkage void syscall_trace(struct pt_regs *regs, int entryexit) { diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig index 1761b74d639..e051cad1f1e 100644 --- a/arch/s390/kvm/Kconfig +++ b/arch/s390/kvm/Kconfig @@ -22,7 +22,6 @@ config KVM select PREEMPT_NOTIFIERS select ANON_INODES select S390_SWITCH_AMODE - select PREEMPT ---help--- Support hosting paravirtualized guest machines using the SIE virtualization capability on the mainframe. This should work diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c index 349581a2610..47a0b642174 100644 --- a/arch/s390/kvm/intercept.c +++ b/arch/s390/kvm/intercept.c @@ -105,6 +105,9 @@ static intercept_handler_t instruction_handlers[256] = { static int handle_noop(struct kvm_vcpu *vcpu) { switch (vcpu->arch.sie_block->icptcode) { + case 0x0: + vcpu->stat.exit_null++; + break; case 0x10: vcpu->stat.exit_external_request++; break; diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 98d1e73e01f..0ac36a649eb 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -31,6 +31,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { "userspace_handled", VCPU_STAT(exit_userspace) }, + { "exit_null", VCPU_STAT(exit_null) }, { "exit_validity", VCPU_STAT(exit_validity) }, { "exit_stop_request", VCPU_STAT(exit_stop_request) }, { "exit_external_request", VCPU_STAT(exit_external_request) }, @@ -221,10 +222,6 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) vcpu->arch.guest_fpregs.fpc &= FPC_VALID_MASK; restore_fp_regs(&vcpu->arch.guest_fpregs); restore_access_regs(vcpu->arch.guest_acrs); - - if (signal_pending(current)) - atomic_set_mask(CPUSTAT_STOP_INT, - &vcpu->arch.sie_block->cpuflags); } void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) diff --git a/arch/s390/mm/Makefile b/arch/s390/mm/Makefile index fb988a48a75..2a745813454 100644 --- a/arch/s390/mm/Makefile +++ b/arch/s390/mm/Makefile @@ -5,3 +5,4 @@ obj-y := init.o fault.o extmem.o mmap.o vmem.o pgtable.o obj-$(CONFIG_CMM) += cmm.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o +obj-$(CONFIG_PAGE_STATES) += page-states.o diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index fa31de6ae97..29f3a63806b 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -126,6 +126,9 @@ void __init mem_init(void) /* clear the zero-page */ memset(empty_zero_page, 0, PAGE_SIZE); + /* Setup guest page hinting */ + cmma_init(); + /* this will put all low memory onto the freelists */ totalram_pages += free_all_bootmem(); diff --git a/arch/s390/mm/page-states.c b/arch/s390/mm/page-states.c new file mode 100644 index 00000000000..fc0ad73ffd9 --- /dev/null +++ b/arch/s390/mm/page-states.c @@ -0,0 +1,79 @@ +/* + * arch/s390/mm/page-states.c + * + * Copyright IBM Corp. 2008 + * + * Guest page hinting for unused pages. + * + * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + */ + +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/types.h> +#include <linux/mm.h> +#include <linux/init.h> + +#define ESSA_SET_STABLE 1 +#define ESSA_SET_UNUSED 2 + +static int cmma_flag; + +static int __init cmma(char *str) +{ + char *parm; + parm = strstrip(str); + if (strcmp(parm, "yes") == 0 || strcmp(parm, "on") == 0) { + cmma_flag = 1; + return 1; + } + cmma_flag = 0; + if (strcmp(parm, "no") == 0 || strcmp(parm, "off") == 0) + return 1; + return 0; +} + +__setup("cmma=", cmma); + +void __init cmma_init(void) +{ + register unsigned long tmp asm("0") = 0; + register int rc asm("1") = -EOPNOTSUPP; + + if (!cmma_flag) + return; + asm volatile( + " .insn rrf,0xb9ab0000,%1,%1,0,0\n" + "0: la %0,0\n" + "1:\n" + EX_TABLE(0b,1b) + : "+&d" (rc), "+&d" (tmp)); + if (rc) + cmma_flag = 0; +} + +void arch_free_page(struct page *page, int order) +{ + int i, rc; + + if (!cmma_flag) + return; + for (i = 0; i < (1 << order); i++) + asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0" + : "=&d" (rc) + : "a" ((page_to_pfn(page) + i) << PAGE_SHIFT), + "i" (ESSA_SET_UNUSED)); +} + +void arch_alloc_page(struct page *page, int order) +{ + int i, rc; + + if (!cmma_flag) + return; + for (i = 0; i < (1 << order); i++) + asm volatile(".insn rrf,0xb9ab0000,%0,%1,%2,0" + : "=&d" (rc) + : "a" ((page_to_pfn(page) + i) << PAGE_SHIFT), + "i" (ESSA_SET_STABLE)); +} diff --git a/arch/sparc/kernel/process.c b/arch/sparc/kernel/process.c index e7f35198ae3..36431f377de 100644 --- a/arch/sparc/kernel/process.c +++ b/arch/sparc/kernel/process.c @@ -419,14 +419,26 @@ asmlinkage int sparc_do_fork(unsigned long clone_flags, unsigned long stack_size) { unsigned long parent_tid_ptr, child_tid_ptr; + unsigned long orig_i1 = regs->u_regs[UREG_I1]; + long ret; parent_tid_ptr = regs->u_regs[UREG_I2]; child_tid_ptr = regs->u_regs[UREG_I4]; - return do_fork(clone_flags, stack_start, - regs, stack_size, - (int __user *) parent_tid_ptr, - (int __user *) child_tid_ptr); + ret = do_fork(clone_flags, stack_start, + regs, stack_size, + (int __user *) parent_tid_ptr, + (int __user *) child_tid_ptr); + + /* If we get an error and potentially restart the system + * call, we're screwed because copy_thread() clobbered + * the parent's %o1. So detect that case and restore it + * here. + */ + if ((unsigned long)ret >= -ERESTART_RESTARTBLOCK) + regs->u_regs[UREG_I1] = orig_i1; + + return ret; } /* Copy a Sparc thread. The fork() return value conventions diff --git a/arch/sparc/kernel/sys_sparc.c b/arch/sparc/kernel/sys_sparc.c index f188b5dc9fd..e995491c443 100644 --- a/arch/sparc/kernel/sys_sparc.c +++ b/arch/sparc/kernel/sys_sparc.c @@ -223,8 +223,7 @@ int sparc_mmap_check(unsigned long addr, unsigned long len, unsigned long flags) { if (ARCH_SUN4C_SUN4 && (len > 0x20000000 || - ((flags & MAP_FIXED) && - addr < 0xe0000000 && addr + len > 0x20000000))) + (addr < 0xe0000000 && addr + len > 0x20000000))) return -EINVAL; /* See asm-sparc/uaccess.h */ diff --git a/arch/sparc64/kernel/process.c b/arch/sparc64/kernel/process.c index 500ac6d483a..4129c044985 100644 --- a/arch/sparc64/kernel/process.c +++ b/arch/sparc64/kernel/process.c @@ -503,6 +503,8 @@ asmlinkage long sparc_do_fork(unsigned long clone_flags, unsigned long stack_size) { int __user *parent_tid_ptr, *child_tid_ptr; + unsigned long orig_i1 = regs->u_regs[UREG_I1]; + long ret; #ifdef CONFIG_COMPAT if (test_thread_flag(TIF_32BIT)) { @@ -515,9 +517,19 @@ asmlinkage long sparc_do_fork(unsigned long clone_flags, child_tid_ptr = (int __user *) regs->u_regs[UREG_I4]; } - return do_fork(clone_flags, stack_start, - regs, stack_size, - parent_tid_ptr, child_tid_ptr); + ret = do_fork(clone_flags, stack_start, + regs, stack_size, + parent_tid_ptr, child_tid_ptr); + + /* If we get an error and potentially restart the system + * call, we're screwed because copy_thread() clobbered + * the parent's %o1. So detect that case and restore it + * here. + */ + if ((unsigned long)ret >= -ERESTART_RESTARTBLOCK) + regs->u_regs[UREG_I1] = orig_i1; + + return ret; } /* Copy a Sparc thread. The fork() return value conventions diff --git a/arch/sparc64/kernel/smp.c b/arch/sparc64/kernel/smp.c index 3aba47624df..0d6403a630a 100644 --- a/arch/sparc64/kernel/smp.c +++ b/arch/sparc64/kernel/smp.c @@ -865,21 +865,14 @@ void smp_call_function_client(int irq, struct pt_regs *regs) void *info = call_data->info; clear_softint(1 << irq); - - irq_enter(); - - if (!call_data->wait) { - /* let initiator proceed after getting data */ - atomic_inc(&call_data->finished); - } - - func(info); - - irq_exit(); - if (call_data->wait) { /* let initiator proceed only after completion */ + func(info); atomic_inc(&call_data->finished); + } else { + /* let initiator proceed after getting data */ + atomic_inc(&call_data->finished); + func(info); } } @@ -1041,9 +1034,7 @@ void smp_receive_signal(int cpu) void smp_receive_signal_client(int irq, struct pt_regs *regs) { - irq_enter(); clear_softint(1 << irq); - irq_exit(); } void smp_new_mmu_context_version_client(int irq, struct pt_regs *regs) @@ -1051,8 +1042,6 @@ void smp_new_mmu_context_version_client(int irq, struct pt_regs *regs) struct mm_struct *mm; unsigned long flags; - irq_enter(); - clear_softint(1 << irq); /* See if we need to allocate a new TLB context because @@ -1072,8 +1061,6 @@ void smp_new_mmu_context_version_client(int irq, struct pt_regs *regs) load_secondary_context(mm); __flush_tlb_mm(CTX_HWBITS(mm->context), SECONDARY_CONTEXT); - - irq_exit(); } void smp_new_mmu_context_version(void) @@ -1239,8 +1226,6 @@ void smp_penguin_jailcell(int irq, struct pt_regs *regs) { clear_softint(1 << irq); - irq_enter(); - preempt_disable(); __asm__ __volatile__("flushw"); @@ -1253,8 +1238,6 @@ void smp_penguin_jailcell(int irq, struct pt_regs *regs) prom_world(0); preempt_enable(); - - irq_exit(); } /* /proc/profile writes can call this, don't __init it please. */ diff --git a/arch/sparc64/kernel/sys_sparc.c b/arch/sparc64/kernel/sys_sparc.c index 8d4761f15fa..0dbc941f130 100644 --- a/arch/sparc64/kernel/sys_sparc.c +++ b/arch/sparc64/kernel/sys_sparc.c @@ -549,13 +549,13 @@ int sparc64_mmap_check(unsigned long addr, unsigned long len, if (len >= STACK_TOP32) return -EINVAL; - if ((flags & MAP_FIXED) && addr > STACK_TOP32 - len) + if (addr > STACK_TOP32 - len) return -EINVAL; } else { if (len >= VA_EXCLUDE_START) return -EINVAL; - if ((flags & MAP_FIXED) && invalid_64bit_range(addr, len)) + if (invalid_64bit_range(addr, len)) return -EINVAL; } diff --git a/arch/sparc64/kernel/sys_sparc32.c b/arch/sparc64/kernel/sys_sparc32.c index 161ce4710fe..1aa4288125f 100644 --- a/arch/sparc64/kernel/sys_sparc32.c +++ b/arch/sparc64/kernel/sys_sparc32.c @@ -236,13 +236,6 @@ asmlinkage long sys32_getegid16(void) /* 32-bit timeval and related flotsam. */ -static long get_tv32(struct timeval *o, struct compat_timeval __user *i) -{ - return (!access_ok(VERIFY_READ, i, sizeof(*i)) || - (__get_user(o->tv_sec, &i->tv_sec) | - __get_user(o->tv_usec, &i->tv_usec))); -} - static inline long put_tv32(struct compat_timeval __user *o, struct timeval *i) { return (!access_ok(VERIFY_WRITE, o, sizeof(*o)) || @@ -757,30 +750,6 @@ asmlinkage long sys32_settimeofday(struct compat_timeval __user *tv, return do_sys_settimeofday(tv ? &kts : NULL, tz ? &ktz : NULL); } -asmlinkage long sys32_utimes(char __user *filename, - struct compat_timeval __user *tvs) -{ - struct timespec tv[2]; - - if (tvs) { - struct timeval ktvs[2]; - if (get_tv32(&ktvs[0], tvs) || - get_tv32(&ktvs[1], 1+tvs)) - return -EFAULT; - - if (ktvs[0].tv_usec < 0 || ktvs[0].tv_usec >= 1000000 || - ktvs[1].tv_usec < 0 || ktvs[1].tv_usec >= 1000000) - return -EINVAL; - - tv[0].tv_sec = ktvs[0].tv_sec; - tv[0].tv_nsec = 1000 * ktvs[0].tv_usec; - tv[1].tv_sec = ktvs[1].tv_sec; - tv[1].tv_nsec = 1000 * ktvs[1].tv_usec; - } - - return do_utimes(AT_FDCWD, filename, tvs ? tv : NULL, 0); -} - /* These are here just in case some old sparc32 binary calls it. */ asmlinkage long sys32_pause(void) { diff --git a/arch/sparc64/kernel/systbls.S b/arch/sparc64/kernel/systbls.S index a4fef2ba1ae..8b5282d433c 100644 --- a/arch/sparc64/kernel/systbls.S +++ b/arch/sparc64/kernel/systbls.S @@ -45,7 +45,7 @@ sys_call_table32: /*120*/ .word compat_sys_readv, compat_sys_writev, sys32_settimeofday, sys32_fchown16, sys_fchmod .word sys_nis_syscall, sys32_setreuid16, sys32_setregid16, sys_rename, sys_truncate /*130*/ .word sys_ftruncate, sys_flock, compat_sys_lstat64, sys_nis_syscall, sys_nis_syscall - .word sys_nis_syscall, sys32_mkdir, sys_rmdir, sys32_utimes, compat_sys_stat64 + .word sys_nis_syscall, sys32_mkdir, sys_rmdir, compat_sys_utimes, compat_sys_stat64 /*140*/ .word sys32_sendfile64, sys_nis_syscall, sys32_futex, sys_gettid, compat_sys_getrlimit .word compat_sys_setrlimit, sys_pivot_root, sys32_prctl, sys_pciconfig_read, sys_pciconfig_write /*150*/ .word sys_nis_syscall, sys_inotify_init, sys_inotify_add_watch, sys_poll, sys_getdents64 diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c index 4cad0b32b0a..ec3e2c72302 100644 --- a/arch/sparc64/mm/init.c +++ b/arch/sparc64/mm/init.c @@ -771,6 +771,9 @@ static void __init find_ramdisk(unsigned long phys_base) initrd_end = ramdisk_image + sparc_ramdisk_size; lmb_reserve(initrd_start, initrd_end); + + initrd_start += PAGE_OFFSET; + initrd_end += PAGE_OFFSET; } #endif } @@ -2362,16 +2365,3 @@ void __flush_tlb_all(void) __asm__ __volatile__("wrpr %0, 0, %%pstate" : : "r" (pstate)); } - -#ifdef CONFIG_MEMORY_HOTPLUG - -void online_page(struct page *page) -{ - ClearPageReserved(page); - init_page_count(page); - __free_page(page); - totalram_pages++; - num_physpages++; -} - -#endif /* CONFIG_MEMORY_HOTPLUG */ diff --git a/arch/um/Makefile b/arch/um/Makefile index 01b97c19a8b..dbeab15e7bb 100644 --- a/arch/um/Makefile +++ b/arch/um/Makefile @@ -77,10 +77,7 @@ include $(srctree)/$(ARCH_DIR)/Makefile-os-$(OS) KERNEL_DEFINES = $(strip -Derrno=kernel_errno -Dsigprocmask=kernel_sigprocmask \ -Dmktime=kernel_mktime $(ARCH_KERNEL_DEFINES)) KBUILD_CFLAGS += $(KERNEL_DEFINES) -# Disable unit-at-a-time mode on pre-gcc-4.0 compilers, it makes gcc use -# a lot more stack due to the lack of sharing of stacklots: -KBUILD_CFLAGS += $(shell if [ $(call cc-version) -lt 0400 ] ; then \ - echo $(call cc-option,-fno-unit-at-a-time); fi ;) +KBUILD_CFLAGS += $(call cc-option,-fno-unit-at-a-time,) PHONY += linux diff --git a/arch/um/drivers/line.c b/arch/um/drivers/line.c index 10b86e1cc65..5047490fc29 100644 --- a/arch/um/drivers/line.c +++ b/arch/um/drivers/line.c @@ -191,9 +191,9 @@ void line_flush_chars(struct tty_struct *tty) line_flush_buffer(tty); } -void line_put_char(struct tty_struct *tty, unsigned char ch) +int line_put_char(struct tty_struct *tty, unsigned char ch) { - line_write(tty, &ch, sizeof(ch)); + return line_write(tty, &ch, sizeof(ch)); } int line_write(struct tty_struct *tty, const unsigned char *buf, int len) diff --git a/arch/um/include/line.h b/arch/um/include/line.h index 1223f2c844b..979b73e6352 100644 --- a/arch/um/include/line.h +++ b/arch/um/include/line.h @@ -71,7 +71,7 @@ extern int line_setup(struct line *lines, unsigned int sizeof_lines, char *init, char **error_out); extern int line_write(struct tty_struct *tty, const unsigned char *buf, int len); -extern void line_put_char(struct tty_struct *tty, unsigned char ch); +extern int line_put_char(struct tty_struct *tty, unsigned char ch); extern void line_set_termios(struct tty_struct *tty, struct ktermios * old); extern int line_chars_in_buffer(struct tty_struct *tty); extern void line_flush_buffer(struct tty_struct *tty); diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c index 9ee007be914..369cf065b6a 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -172,10 +172,3 @@ void reserve_top_address(unsigned long reserve) __FIXADDR_TOP = -reserve - PAGE_SIZE; __VMALLOC_RESERVE += reserve; } - -int pmd_bad(pmd_t pmd) -{ - WARN_ON_ONCE(pmd_bad_v1(pmd) != pmd_bad_v2(pmd)); - - return pmd_bad_v1(pmd); -} diff --git a/block/blk-core.c b/block/blk-core.c index b754a4a2f9b..2987fe47b5e 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -54,15 +54,16 @@ static DEFINE_PER_CPU(struct list_head, blk_cpu_done); static void drive_stat_acct(struct request *rq, int new_io) { + struct hd_struct *part; int rw = rq_data_dir(rq); if (!blk_fs_request(rq) || !rq->rq_disk) return; - if (!new_io) { - __all_stat_inc(rq->rq_disk, merges[rw], rq->sector); - } else { - struct hd_struct *part = get_part(rq->rq_disk, rq->sector); + part = get_part(rq->rq_disk, rq->sector); + if (!new_io) + __all_stat_inc(rq->rq_disk, part, merges[rw], rq->sector); + else { disk_round_stats(rq->rq_disk); rq->rq_disk->in_flight++; if (part) { @@ -253,9 +254,11 @@ EXPORT_SYMBOL(__generic_unplug_device); **/ void generic_unplug_device(struct request_queue *q) { - spin_lock_irq(q->queue_lock); - __generic_unplug_device(q); - spin_unlock_irq(q->queue_lock); + if (blk_queue_plugged(q)) { + spin_lock_irq(q->queue_lock); + __generic_unplug_device(q); + spin_unlock_irq(q->queue_lock); + } } EXPORT_SYMBOL(generic_unplug_device); @@ -1536,10 +1539,11 @@ static int __end_that_request_first(struct request *req, int error, } if (blk_fs_request(req) && req->rq_disk) { + struct hd_struct *part = get_part(req->rq_disk, req->sector); const int rw = rq_data_dir(req); - all_stat_add(req->rq_disk, sectors[rw], - nr_bytes >> 9, req->sector); + all_stat_add(req->rq_disk, part, sectors[rw], + nr_bytes >> 9, req->sector); } total_bytes = bio_nbytes = 0; @@ -1725,8 +1729,8 @@ static void end_that_request_last(struct request *req, int error) const int rw = rq_data_dir(req); struct hd_struct *part = get_part(disk, req->sector); - __all_stat_inc(disk, ios[rw], req->sector); - __all_stat_add(disk, ticks[rw], duration, req->sector); + __all_stat_inc(disk, part, ios[rw], req->sector); + __all_stat_add(disk, part, ticks[rw], duration, req->sector); disk_round_stats(disk); disk->in_flight--; if (part) { diff --git a/block/blk-ioc.c b/block/blk-ioc.c index e34df7c9fc3..012f065ac8e 100644 --- a/block/blk-ioc.c +++ b/block/blk-ioc.c @@ -41,8 +41,8 @@ int put_io_context(struct io_context *ioc) rcu_read_lock(); if (ioc->aic && ioc->aic->dtor) ioc->aic->dtor(ioc->aic); - rcu_read_unlock(); cfq_dtor(ioc); + rcu_read_unlock(); kmem_cache_free(iocontext_cachep, ioc); return 1; diff --git a/block/blk-merge.c b/block/blk-merge.c index 73b23562af2..651136aae76 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -149,9 +149,9 @@ static int blk_phys_contig_segment(struct request_queue *q, struct bio *bio, static int blk_hw_contig_segment(struct request_queue *q, struct bio *bio, struct bio *nxt) { - if (unlikely(!bio_flagged(bio, BIO_SEG_VALID))) + if (!bio_flagged(bio, BIO_SEG_VALID)) blk_recount_segments(q, bio); - if (unlikely(!bio_flagged(nxt, BIO_SEG_VALID))) + if (!bio_flagged(nxt, BIO_SEG_VALID)) blk_recount_segments(q, nxt); if (!BIOVEC_VIRT_MERGEABLE(__BVEC_END(bio), __BVEC_START(nxt)) || BIOVEC_VIRT_OVERSIZE(bio->bi_hw_back_size + nxt->bi_hw_front_size)) @@ -312,9 +312,9 @@ int ll_back_merge_fn(struct request_queue *q, struct request *req, q->last_merge = NULL; return 0; } - if (unlikely(!bio_flagged(req->biotail, BIO_SEG_VALID))) + if (!bio_flagged(req->biotail, BIO_SEG_VALID)) blk_recount_segments(q, req->biotail); - if (unlikely(!bio_flagged(bio, BIO_SEG_VALID))) + if (!bio_flagged(bio, BIO_SEG_VALID)) blk_recount_segments(q, bio); len = req->biotail->bi_hw_back_size + bio->bi_hw_front_size; if (BIOVEC_VIRT_MERGEABLE(__BVEC_END(req->biotail), __BVEC_START(bio)) @@ -352,9 +352,9 @@ int ll_front_merge_fn(struct request_queue *q, struct request *req, return 0; } len = bio->bi_hw_back_size + req->bio->bi_hw_front_size; - if (unlikely(!bio_flagged(bio, BIO_SEG_VALID))) + if (!bio_flagged(bio, BIO_SEG_VALID)) blk_recount_segments(q, bio); - if (unlikely(!bio_flagged(req->bio, BIO_SEG_VALID))) + if (!bio_flagged(req->bio, BIO_SEG_VALID)) blk_recount_segments(q, req->bio); if (BIOVEC_VIRT_MERGEABLE(__BVEC_END(bio), __BVEC_START(req->bio)) && !BIOVEC_VIRT_OVERSIZE(len)) { diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index e85c4013e8a..304ec73ab82 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -146,11 +146,13 @@ static ssize_t queue_nomerges_store(struct request_queue *q, const char *page, unsigned long nm; ssize_t ret = queue_var_store(&nm, page, count); + spin_lock_irq(q->queue_lock); if (nm) - set_bit(QUEUE_FLAG_NOMERGES, &q->queue_flags); + queue_flag_set(QUEUE_FLAG_NOMERGES, q); else - clear_bit(QUEUE_FLAG_NOMERGES, &q->queue_flags); + queue_flag_clear(QUEUE_FLAG_NOMERGES, q); + spin_unlock_irq(q->queue_lock); return ret; } diff --git a/block/blk-tag.c b/block/blk-tag.c index de64e042997..32667beb03e 100644 --- a/block/blk-tag.c +++ b/block/blk-tag.c @@ -70,7 +70,7 @@ void __blk_queue_free_tags(struct request_queue *q) __blk_free_tags(bqt); q->queue_tags = NULL; - queue_flag_clear(QUEUE_FLAG_QUEUED, q); + queue_flag_clear_unlocked(QUEUE_FLAG_QUEUED, q); } /** @@ -98,7 +98,7 @@ EXPORT_SYMBOL(blk_free_tags); **/ void blk_queue_free_tags(struct request_queue *q) { - queue_flag_clear(QUEUE_FLAG_QUEUED, q); + queue_flag_clear_unlocked(QUEUE_FLAG_QUEUED, q); } EXPORT_SYMBOL(blk_queue_free_tags); @@ -171,6 +171,9 @@ EXPORT_SYMBOL(blk_init_tags); * @q: the request queue for the device * @depth: the maximum queue depth supported * @tags: the tag to use + * + * Queue lock must be held here if the function is called to resize an + * existing map. **/ int blk_queue_init_tags(struct request_queue *q, int depth, struct blk_queue_tag *tags) @@ -197,7 +200,7 @@ int blk_queue_init_tags(struct request_queue *q, int depth, * assign it, all done */ q->queue_tags = tags; - queue_flag_set(QUEUE_FLAG_QUEUED, q); + queue_flag_set_unlocked(QUEUE_FLAG_QUEUED, q); INIT_LIST_HEAD(&q->tag_busy_list); return 0; fail: diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index f4e1006c253..b399c62936e 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1142,6 +1142,17 @@ static void cfq_put_queue(struct cfq_queue *cfqq) kmem_cache_free(cfq_pool, cfqq); } +static void +__call_for_each_cic(struct io_context *ioc, + void (*func)(struct io_context *, struct cfq_io_context *)) +{ + struct cfq_io_context *cic; + struct hlist_node *n; + + hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list) + func(ioc, cic); +} + /* * Call func for each cic attached to this ioc. */ @@ -1149,12 +1160,8 @@ static void call_for_each_cic(struct io_context *ioc, void (*func)(struct io_context *, struct cfq_io_context *)) { - struct cfq_io_context *cic; - struct hlist_node *n; - rcu_read_lock(); - hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list) - func(ioc, cic); + __call_for_each_cic(ioc, func); rcu_read_unlock(); } @@ -1198,7 +1205,7 @@ static void cfq_free_io_context(struct io_context *ioc) * should be ok to iterate over the known list, we will see all cic's * since no new ones are added. */ - call_for_each_cic(ioc, cic_free_func); + __call_for_each_cic(ioc, cic_free_func); } static void cfq_exit_cfqq(struct cfq_data *cfqd, struct cfq_queue *cfqq) @@ -1296,10 +1303,10 @@ static void cfq_init_prio_data(struct cfq_queue *cfqq, struct io_context *ioc) printk(KERN_ERR "cfq: bad prio %x\n", ioprio_class); case IOPRIO_CLASS_NONE: /* - * no prio set, place us in the middle of the BE classes + * no prio set, inherit CPU scheduling settings */ cfqq->ioprio = task_nice_ioprio(tsk); - cfqq->ioprio_class = IOPRIO_CLASS_BE; + cfqq->ioprio_class = task_nice_ioclass(tsk); break; case IOPRIO_CLASS_RT: cfqq->ioprio = task_ioprio(ioc); diff --git a/drivers/accessibility/Kconfig b/drivers/accessibility/Kconfig index 1264c4b9809..ef3b65bfdd0 100644 --- a/drivers/accessibility/Kconfig +++ b/drivers/accessibility/Kconfig @@ -1,7 +1,17 @@ menuconfig ACCESSIBILITY bool "Accessibility support" ---help--- - Enable a submenu where accessibility items may be enabled. + Accessibility handles all special kinds of hardware devices or + software adapters which help people with disabilities (e.g. + blindness) to use computers. + + That includes braille devices, speech synthesis, keyboard + remapping, etc. + + Say Y here to get to see options for accessibility. + This option alone does not add any kernel code. + + If you say N, all options in this submenu will be skipped and disabled. If unsure, say N. diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig index 1c11df9a5f3..9bf2986a278 100644 --- a/drivers/ata/Kconfig +++ b/drivers/ata/Kconfig @@ -205,8 +205,8 @@ config SATA_VITESSE If unsure, say N. config SATA_INIC162X - tristate "Initio 162x SATA support (HIGHLY EXPERIMENTAL)" - depends on PCI && EXPERIMENTAL + tristate "Initio 162x SATA support" + depends on PCI help This option enables support for Initio 162x Serial ATA. @@ -697,6 +697,15 @@ config PATA_SCC If unsure, say N. +config PATA_SCH + tristate "Intel SCH PATA support" + depends on PCI + help + This option enables support for Intel SCH PATA on the Intel + SCH (US15W, US15L, UL11L) series host controllers. + + If unsure, say N. + config PATA_BF54X tristate "Blackfin 54x ATAPI support" depends on BF542 || BF548 || BF549 diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile index b693d829383..674965fa326 100644 --- a/drivers/ata/Makefile +++ b/drivers/ata/Makefile @@ -67,6 +67,7 @@ obj-$(CONFIG_PATA_SIS) += pata_sis.o obj-$(CONFIG_PATA_TRIFLEX) += pata_triflex.o obj-$(CONFIG_PATA_IXP4XX_CF) += pata_ixp4xx_cf.o obj-$(CONFIG_PATA_SCC) += pata_scc.o +obj-$(CONFIG_PATA_SCH) += pata_sch.o obj-$(CONFIG_PATA_BF54X) += pata_bf54x.o obj-$(CONFIG_PATA_PLATFORM) += pata_platform.o obj-$(CONFIG_PATA_OF_PLATFORM) += pata_of_platform.o diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 8cace9aa9c0..97f83fb2ee2 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -1267,9 +1267,7 @@ static int ahci_check_ready(struct ata_link *link) void __iomem *port_mmio = ahci_port_base(link->ap); u8 status = readl(port_mmio + PORT_TFDATA) & 0xFF; - if (!(status & ATA_BUSY)) - return 1; - return 0; + return ata_check_ready(status); } static int ahci_softreset(struct ata_link *link, unsigned int *class, diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c index 47aeccd52fa..75a406f5e69 100644 --- a/drivers/ata/ata_generic.c +++ b/drivers/ata/ata_generic.c @@ -152,6 +152,12 @@ static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id if (dev->vendor == PCI_VENDOR_ID_AL) ata_pci_bmdma_clear_simplex(dev); + if (dev->vendor == PCI_VENDOR_ID_ATI) { + int rc = pcim_enable_device(dev); + if (rc < 0) + return rc; + pcim_pin_device(dev); + } return ata_pci_sff_init_one(dev, ppi, &generic_sht, NULL); } diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c index ea2c7649d39..a9027b8fbdd 100644 --- a/drivers/ata/ata_piix.c +++ b/drivers/ata/ata_piix.c @@ -1348,6 +1348,8 @@ static void __devinit piix_init_sidpr(struct ata_host *host) { struct pci_dev *pdev = to_pci_dev(host->dev); struct piix_host_priv *hpriv = host->private_data; + struct ata_device *dev0 = &host->ports[0]->link.device[0]; + u32 scontrol; int i; /* check for availability */ @@ -1366,6 +1368,29 @@ static void __devinit piix_init_sidpr(struct ata_host *host) return; hpriv->sidpr = pcim_iomap_table(pdev)[PIIX_SIDPR_BAR]; + + /* SCR access via SIDPR doesn't work on some configurations. + * Give it a test drive by inhibiting power save modes which + * we'll do anyway. + */ + scontrol = piix_sidpr_read(dev0, SCR_CONTROL); + + /* if IPM is already 3, SCR access is probably working. Don't + * un-inhibit power save modes as BIOS might have inhibited + * them for a reason. + */ + if ((scontrol & 0xf00) != 0x300) { + scontrol |= 0x300; + piix_sidpr_write(dev0, SCR_CONTROL, scontrol); + scontrol = piix_sidpr_read(dev0, SCR_CONTROL); + + if ((scontrol & 0xf00) != 0x300) { + dev_printk(KERN_INFO, host->dev, "SCR access via " + "SIDPR is available but doesn't work\n"); + return; + } + } + host->ports[0]->ops = &piix_sidpr_sata_ops; host->ports[1]->ops = &piix_sidpr_sata_ops; } diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 3bc48853820..927b692d723 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -6292,6 +6292,7 @@ EXPORT_SYMBOL_GPL(ata_eh_freeze_port); EXPORT_SYMBOL_GPL(ata_eh_thaw_port); EXPORT_SYMBOL_GPL(ata_eh_qc_complete); EXPORT_SYMBOL_GPL(ata_eh_qc_retry); +EXPORT_SYMBOL_GPL(ata_eh_analyze_ncq_error); EXPORT_SYMBOL_GPL(ata_do_eh); EXPORT_SYMBOL_GPL(ata_std_error_handler); diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index 61dcd0026c6..62e033146be 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -1357,7 +1357,7 @@ static void ata_eh_analyze_serror(struct ata_link *link) * LOCKING: * Kernel thread context (may sleep). */ -static void ata_eh_analyze_ncq_error(struct ata_link *link) +void ata_eh_analyze_ncq_error(struct ata_link *link) { struct ata_port *ap = link->ap; struct ata_eh_context *ehc = &link->eh_context; diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c index 2ec65a8fda7..3c2d2289f85 100644 --- a/drivers/ata/libata-sff.c +++ b/drivers/ata/libata-sff.c @@ -314,11 +314,7 @@ static int ata_sff_check_ready(struct ata_link *link) { u8 status = link->ap->ops->sff_check_status(link->ap); - if (!(status & ATA_BUSY)) - return 1; - if (status == 0xff) - return -ENODEV; - return 0; + return ata_check_ready(status); } /** diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c index c5f91e62994..fbe60571155 100644 --- a/drivers/ata/pata_acpi.c +++ b/drivers/ata/pata_acpi.c @@ -259,6 +259,12 @@ static int pacpi_init_one (struct pci_dev *pdev, const struct pci_device_id *id) .port_ops = &pacpi_ops, }; const struct ata_port_info *ppi[] = { &info, NULL }; + if (pdev->vendor == PCI_VENDOR_ID_ATI) { + int rc = pcim_enable_device(pdev); + if (rc < 0) + return rc; + pcim_pin_device(pdev); + } return ata_pci_sff_init_one(pdev, ppi, &pacpi_sht, NULL); } diff --git a/drivers/ata/pata_sch.c b/drivers/ata/pata_sch.c new file mode 100644 index 00000000000..c8cc027789f --- /dev/null +++ b/drivers/ata/pata_sch.c @@ -0,0 +1,206 @@ +/* + * pata_sch.c - Intel SCH PATA controllers + * + * Copyright (c) 2008 Alek Du <alek.du@intel.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License 2 as published + * by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * + */ + +/* + * Supports: + * Intel SCH (AF82US15W, AF82US15L, AF82UL11L) chipsets -- see spec at: + * http://download.intel.com/design/chipsets/embedded/datashts/319537.pdf + */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/pci.h> +#include <linux/init.h> +#include <linux/blkdev.h> +#include <linux/delay.h> +#include <linux/device.h> +#include <scsi/scsi_host.h> +#include <linux/libata.h> +#include <linux/dmi.h> + +#define DRV_NAME "pata_sch" +#define DRV_VERSION "0.2" + +/* see SCH datasheet page 351 */ +enum { + D0TIM = 0x80, /* Device 0 Timing Register */ + D1TIM = 0x84, /* Device 1 Timing Register */ + PM = 0x07, /* PIO Mode Bit Mask */ + MDM = (0x03 << 8), /* Multi-word DMA Mode Bit Mask */ + UDM = (0x07 << 16), /* Ultra DMA Mode Bit Mask */ + PPE = (1 << 30), /* Prefetch/Post Enable */ + USD = (1 << 31), /* Use Synchronous DMA */ +}; + +static int sch_init_one(struct pci_dev *pdev, + const struct pci_device_id *ent); +static void sch_set_piomode(struct ata_port *ap, struct ata_device *adev); +static void sch_set_dmamode(struct ata_port *ap, struct ata_device *adev); + +static const struct pci_device_id sch_pci_tbl[] = { + /* Intel SCH PATA Controller */ + { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_SCH_IDE), 0 }, + { } /* terminate list */ +}; + +static struct pci_driver sch_pci_driver = { + .name = DRV_NAME, + .id_table = sch_pci_tbl, + .probe = sch_init_one, + .remove = ata_pci_remove_one, +#ifdef CONFIG_PM + .suspend = ata_pci_device_suspend, + .resume = ata_pci_device_resume, +#endif +}; + +static struct scsi_host_template sch_sht = { + ATA_BMDMA_SHT(DRV_NAME), +}; + +static struct ata_port_operations sch_pata_ops = { + .inherits = &ata_bmdma_port_ops, + .cable_detect = ata_cable_unknown, + .set_piomode = sch_set_piomode, + .set_dmamode = sch_set_dmamode, +}; + +static struct ata_port_info sch_port_info = { + .flags = 0, + .pio_mask = ATA_PIO4, /* pio0-4 */ + .mwdma_mask = ATA_MWDMA2, /* mwdma0-2 */ + .udma_mask = ATA_UDMA5, /* udma0-5 */ + .port_ops = &sch_pata_ops, +}; + +MODULE_AUTHOR("Alek Du <alek.du@intel.com>"); +MODULE_DESCRIPTION("SCSI low-level driver for Intel SCH PATA controllers"); +MODULE_LICENSE("GPL"); +MODULE_DEVICE_TABLE(pci, sch_pci_tbl); +MODULE_VERSION(DRV_VERSION); + +/** + * sch_set_piomode - Initialize host controller PATA PIO timings + * @ap: Port whose timings we are configuring + * @adev: ATA device + * + * Set PIO mode for device, in host controller PCI config space. + * + * LOCKING: + * None (inherited from caller). + */ + +static void sch_set_piomode(struct ata_port *ap, struct ata_device *adev) +{ + unsigned int pio = adev->pio_mode - XFER_PIO_0; + struct pci_dev *dev = to_pci_dev(ap->host->dev); + unsigned int port = adev->devno ? D1TIM : D0TIM; + unsigned int data; + + pci_read_config_dword(dev, port, &data); + /* see SCH datasheet page 351 */ + /* set PIO mode */ + data &= ~(PM | PPE); + data |= pio; + /* enable PPE for block device */ + if (adev->class == ATA_DEV_ATA) + data |= PPE; + pci_write_config_dword(dev, port, data); +} + +/** + * sch_set_dmamode - Initialize host controller PATA DMA timings + * @ap: Port whose timings we are configuring + * @adev: ATA device + * + * Set MW/UDMA mode for device, in host controller PCI config space. + * + * LOCKING: + * None (inherited from caller). + */ + +static void sch_set_dmamode(struct ata_port *ap, struct ata_device *adev) +{ + unsigned int dma_mode = adev->dma_mode; + struct pci_dev *dev = to_pci_dev(ap->host->dev); + unsigned int port = adev->devno ? D1TIM : D0TIM; + unsigned int data; + + pci_read_config_dword(dev, port, &data); + /* see SCH datasheet page 351 */ + if (dma_mode >= XFER_UDMA_0) { + /* enable Synchronous DMA mode */ + data |= USD; + data &= ~UDM; + data |= (dma_mode - XFER_UDMA_0) << 16; + } else { /* must be MWDMA mode, since we masked SWDMA already */ + data &= ~(USD | MDM); + data |= (dma_mode - XFER_MW_DMA_0) << 8; + } + pci_write_config_dword(dev, port, data); +} + +/** + * sch_init_one - Register SCH ATA PCI device with kernel services + * @pdev: PCI device to register + * @ent: Entry in sch_pci_tbl matching with @pdev + * + * LOCKING: + * Inherited from PCI layer (may sleep). + * + * RETURNS: + * Zero on success, or -ERRNO value. + */ + +static int __devinit sch_init_one(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + static int printed_version; + const struct ata_port_info *ppi[] = { &sch_port_info, NULL }; + struct ata_host *host; + int rc; + + if (!printed_version++) + dev_printk(KERN_DEBUG, &pdev->dev, + "version " DRV_VERSION "\n"); + + /* enable device and prepare host */ + rc = pcim_enable_device(pdev); + if (rc) + return rc; + rc = ata_pci_sff_prepare_host(pdev, ppi, &host); + if (rc) + return rc; + pci_set_master(pdev); + return ata_pci_sff_activate_host(host, ata_sff_interrupt, &sch_sht); +} + +static int __init sch_init(void) +{ + return pci_register_driver(&sch_pci_driver); +} + +static void __exit sch_exit(void) +{ + pci_unregister_driver(&sch_pci_driver); +} + +module_init(sch_init); +module_exit(sch_exit); diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c index d27bb9a2568..3ead02fe379 100644 --- a/drivers/ata/sata_inic162x.c +++ b/drivers/ata/sata_inic162x.c @@ -10,13 +10,33 @@ * right. Documentation is available at initio's website but it only * documents registers (not programming model). * - * - ATA disks work. - * - Hotplug works. - * - ATAPI read works but burning doesn't. This thing is really - * peculiar about ATAPI and I couldn't figure out how ATAPI PIO and - * ATAPI DMA WRITE should be programmed. If you've got a clue, be - * my guest. - * - Both STR and STD work. + * This driver has interesting history. The first version was written + * from the documentation and a 2.4 IDE driver posted on a Taiwan + * company, which didn't use any IDMA features and couldn't handle + * LBA48. The resulting driver couldn't handle LBA48 devices either + * making it pretty useless. + * + * After a while, initio picked the driver up, renamed it to + * sata_initio162x, updated it to use IDMA for ATA DMA commands and + * posted it on their website. It only used ATA_PROT_DMA for IDMA and + * attaching both devices and issuing IDMA and !IDMA commands + * simultaneously broke it due to PIRQ masking interaction but it did + * show how to use the IDMA (ADMA + some initio specific twists) + * engine. + * + * Then, I picked up their changes again and here's the usable driver + * which uses IDMA for everything. Everything works now including + * LBA48, CD/DVD burning, suspend/resume and hotplug. There are some + * issues tho. Result Tf is not resported properly, NCQ isn't + * supported yet and CD/DVD writing works with DMA assisted PIO + * protocol (which, for native SATA devices, shouldn't cause any + * noticeable difference). + * + * Anyways, so, here's finally a working driver for inic162x. Enjoy! + * + * initio: If you guys wanna improve the driver regarding result TF + * access and other stuff, please feel free to contact me. I'll be + * happy to assist. */ #include <linux/kernel.h> @@ -28,13 +48,19 @@ #include <scsi/scsi_device.h> #define DRV_NAME "sata_inic162x" -#define DRV_VERSION "0.3" +#define DRV_VERSION "0.4" enum { - MMIO_BAR = 5, + MMIO_BAR_PCI = 5, + MMIO_BAR_CARDBUS = 1, NR_PORTS = 2, + IDMA_CPB_TBL_SIZE = 4 * 32, + + INIC_DMA_BOUNDARY = 0xffffff, + + HOST_ACTRL = 0x08, HOST_CTL = 0x7c, HOST_STAT = 0x7e, HOST_IRQ_STAT = 0xbc, @@ -43,22 +69,37 @@ enum { PORT_SIZE = 0x40, /* registers for ATA TF operation */ - PORT_TF = 0x00, - PORT_ALT_STAT = 0x08, + PORT_TF_DATA = 0x00, + PORT_TF_FEATURE = 0x01, + PORT_TF_NSECT = 0x02, + PORT_TF_LBAL = 0x03, + PORT_TF_LBAM = 0x04, + PORT_TF_LBAH = 0x05, + PORT_TF_DEVICE = 0x06, + PORT_TF_COMMAND = 0x07, + PORT_TF_ALT_STAT = 0x08, PORT_IRQ_STAT = 0x09, PORT_IRQ_MASK = 0x0a, PORT_PRD_CTL = 0x0b, PORT_PRD_ADDR = 0x0c, PORT_PRD_XFERLEN = 0x10, + PORT_CPB_CPBLAR = 0x18, + PORT_CPB_PTQFIFO = 0x1c, /* IDMA register */ PORT_IDMA_CTL = 0x14, + PORT_IDMA_STAT = 0x16, + + PORT_RPQ_FIFO = 0x1e, + PORT_RPQ_CNT = 0x1f, PORT_SCR = 0x20, /* HOST_CTL bits */ HCTL_IRQOFF = (1 << 8), /* global IRQ off */ - HCTL_PWRDWN = (1 << 13), /* power down PHYs */ + HCTL_FTHD0 = (1 << 10), /* fifo threshold 0 */ + HCTL_FTHD1 = (1 << 11), /* fifo threshold 1*/ + HCTL_PWRDWN = (1 << 12), /* power down PHYs */ HCTL_SOFTRST = (1 << 13), /* global reset (no phy reset) */ HCTL_RPGSEL = (1 << 15), /* register page select */ @@ -81,9 +122,7 @@ enum { PIRQ_PENDING = (1 << 7), /* port IRQ pending (STAT only) */ PIRQ_ERR = PIRQ_OFFLINE | PIRQ_ONLINE | PIRQ_FATAL, - - PIRQ_MASK_DMA_READ = PIRQ_REPLY | PIRQ_ATA, - PIRQ_MASK_OTHER = PIRQ_REPLY | PIRQ_COMPLETE, + PIRQ_MASK_DEFAULT = PIRQ_REPLY | PIRQ_ATA, PIRQ_MASK_FREEZE = 0xff, /* PORT_PRD_CTL bits */ @@ -96,20 +135,104 @@ enum { IDMA_CTL_RST_IDMA = (1 << 5), /* reset IDMA machinary */ IDMA_CTL_GO = (1 << 7), /* IDMA mode go */ IDMA_CTL_ATA_NIEN = (1 << 8), /* ATA IRQ disable */ + + /* PORT_IDMA_STAT bits */ + IDMA_STAT_PERR = (1 << 0), /* PCI ERROR MODE */ + IDMA_STAT_CPBERR = (1 << 1), /* ADMA CPB error */ + IDMA_STAT_LGCY = (1 << 3), /* ADMA legacy */ + IDMA_STAT_UIRQ = (1 << 4), /* ADMA unsolicited irq */ + IDMA_STAT_STPD = (1 << 5), /* ADMA stopped */ + IDMA_STAT_PSD = (1 << 6), /* ADMA pause */ + IDMA_STAT_DONE = (1 << 7), /* ADMA done */ + + IDMA_STAT_ERR = IDMA_STAT_PERR | IDMA_STAT_CPBERR, + + /* CPB Control Flags*/ + CPB_CTL_VALID = (1 << 0), /* CPB valid */ + CPB_CTL_QUEUED = (1 << 1), /* queued command */ + CPB_CTL_DATA = (1 << 2), /* data, rsvd in datasheet */ + CPB_CTL_IEN = (1 << 3), /* PCI interrupt enable */ + CPB_CTL_DEVDIR = (1 << 4), /* device direction control */ + + /* CPB Response Flags */ + CPB_RESP_DONE = (1 << 0), /* ATA command complete */ + CPB_RESP_REL = (1 << 1), /* ATA release */ + CPB_RESP_IGNORED = (1 << 2), /* CPB ignored */ + CPB_RESP_ATA_ERR = (1 << 3), /* ATA command error */ + CPB_RESP_SPURIOUS = (1 << 4), /* ATA spurious interrupt error */ + CPB_RESP_UNDERFLOW = (1 << 5), /* APRD deficiency length error */ + CPB_RESP_OVERFLOW = (1 << 6), /* APRD exccess length error */ + CPB_RESP_CPB_ERR = (1 << 7), /* CPB error flag */ + + /* PRD Control Flags */ + PRD_DRAIN = (1 << 1), /* ignore data excess */ + PRD_CDB = (1 << 2), /* atapi packet command pointer */ + PRD_DIRECT_INTR = (1 << 3), /* direct interrupt */ + PRD_DMA = (1 << 4), /* data transfer method */ + PRD_WRITE = (1 << 5), /* data dir, rsvd in datasheet */ + PRD_IOM = (1 << 6), /* io/memory transfer */ + PRD_END = (1 << 7), /* APRD chain end */ }; +/* Comman Parameter Block */ +struct inic_cpb { + u8 resp_flags; /* Response Flags */ + u8 error; /* ATA Error */ + u8 status; /* ATA Status */ + u8 ctl_flags; /* Control Flags */ + __le32 len; /* Total Transfer Length */ + __le32 prd; /* First PRD pointer */ + u8 rsvd[4]; + /* 16 bytes */ + u8 feature; /* ATA Feature */ + u8 hob_feature; /* ATA Ex. Feature */ + u8 device; /* ATA Device/Head */ + u8 mirctl; /* Mirror Control */ + u8 nsect; /* ATA Sector Count */ + u8 hob_nsect; /* ATA Ex. Sector Count */ + u8 lbal; /* ATA Sector Number */ + u8 hob_lbal; /* ATA Ex. Sector Number */ + u8 lbam; /* ATA Cylinder Low */ + u8 hob_lbam; /* ATA Ex. Cylinder Low */ + u8 lbah; /* ATA Cylinder High */ + u8 hob_lbah; /* ATA Ex. Cylinder High */ + u8 command; /* ATA Command */ + u8 ctl; /* ATA Control */ + u8 slave_error; /* Slave ATA Error */ + u8 slave_status; /* Slave ATA Status */ + /* 32 bytes */ +} __packed; + +/* Physical Region Descriptor */ +struct inic_prd { + __le32 mad; /* Physical Memory Address */ + __le16 len; /* Transfer Length */ + u8 rsvd; + u8 flags; /* Control Flags */ +} __packed; + +struct inic_pkt { + struct inic_cpb cpb; + struct inic_prd prd[LIBATA_MAX_PRD + 1]; /* + 1 for cdb */ + u8 cdb[ATAPI_CDB_LEN]; +} __packed; + struct inic_host_priv { - u16 cached_hctl; + void __iomem *mmio_base; + u16 cached_hctl; }; struct inic_port_priv { - u8 dfl_prdctl; - u8 cached_prdctl; - u8 cached_pirq_mask; + struct inic_pkt *pkt; + dma_addr_t pkt_dma; + u32 *cpb_tbl; + dma_addr_t cpb_tbl_dma; }; static struct scsi_host_template inic_sht = { - ATA_BMDMA_SHT(DRV_NAME), + ATA_BASE_SHT(DRV_NAME), + .sg_tablesize = LIBATA_MAX_PRD, /* maybe it can be larger? */ + .dma_boundary = INIC_DMA_BOUNDARY, }; static const int scr_map[] = { @@ -120,54 +243,34 @@ static const int scr_map[] = { static void __iomem *inic_port_base(struct ata_port *ap) { - return ap->host->iomap[MMIO_BAR] + ap->port_no * PORT_SIZE; -} - -static void __inic_set_pirq_mask(struct ata_port *ap, u8 mask) -{ - void __iomem *port_base = inic_port_base(ap); - struct inic_port_priv *pp = ap->private_data; + struct inic_host_priv *hpriv = ap->host->private_data; - writeb(mask, port_base + PORT_IRQ_MASK); - pp->cached_pirq_mask = mask; -} - -static void inic_set_pirq_mask(struct ata_port *ap, u8 mask) -{ - struct inic_port_priv *pp = ap->private_data; - - if (pp->cached_pirq_mask != mask) - __inic_set_pirq_mask(ap, mask); + return hpriv->mmio_base + ap->port_no * PORT_SIZE; } static void inic_reset_port(void __iomem *port_base) { void __iomem *idma_ctl = port_base + PORT_IDMA_CTL; - u16 ctl; - ctl = readw(idma_ctl); - ctl &= ~(IDMA_CTL_RST_IDMA | IDMA_CTL_ATA_NIEN | IDMA_CTL_GO); + /* stop IDMA engine */ + readw(idma_ctl); /* flush */ + msleep(1); /* mask IRQ and assert reset */ - writew(ctl | IDMA_CTL_RST_IDMA | IDMA_CTL_ATA_NIEN, idma_ctl); + writew(IDMA_CTL_RST_IDMA, idma_ctl); readw(idma_ctl); /* flush */ - - /* give it some time */ msleep(1); /* release reset */ - writew(ctl | IDMA_CTL_ATA_NIEN, idma_ctl); + writew(0, idma_ctl); /* clear irq */ writeb(0xff, port_base + PORT_IRQ_STAT); - - /* reenable ATA IRQ, turn off IDMA mode */ - writew(ctl, idma_ctl); } static int inic_scr_read(struct ata_port *ap, unsigned sc_reg, u32 *val) { - void __iomem *scr_addr = ap->ioaddr.scr_addr; + void __iomem *scr_addr = inic_port_base(ap) + PORT_SCR; void __iomem *addr; if (unlikely(sc_reg >= ARRAY_SIZE(scr_map))) @@ -184,120 +287,126 @@ static int inic_scr_read(struct ata_port *ap, unsigned sc_reg, u32 *val) static int inic_scr_write(struct ata_port *ap, unsigned sc_reg, u32 val) { - void __iomem *scr_addr = ap->ioaddr.scr_addr; - void __iomem *addr; + void __iomem *scr_addr = inic_port_base(ap) + PORT_SCR; if (unlikely(sc_reg >= ARRAY_SIZE(scr_map))) return -EINVAL; - addr = scr_addr + scr_map[sc_reg] * 4; writel(val, scr_addr + scr_map[sc_reg] * 4); return 0; } -/* - * In TF mode, inic162x is very similar to SFF device. TF registers - * function the same. DMA engine behaves similary using the same PRD - * format as BMDMA but different command register, interrupt and event - * notification methods are used. The following inic_bmdma_*() - * functions do the impedance matching. - */ -static void inic_bmdma_setup(struct ata_queued_cmd *qc) +static void inic_stop_idma(struct ata_port *ap) { - struct ata_port *ap = qc->ap; - struct inic_port_priv *pp = ap->private_data; void __iomem *port_base = inic_port_base(ap); - int rw = qc->tf.flags & ATA_TFLAG_WRITE; - - /* make sure device sees PRD table writes */ - wmb(); - - /* load transfer length */ - writel(qc->nbytes, port_base + PORT_PRD_XFERLEN); - - /* turn on DMA and specify data direction */ - pp->cached_prdctl = pp->dfl_prdctl | PRD_CTL_DMAEN; - if (!rw) - pp->cached_prdctl |= PRD_CTL_WR; - writeb(pp->cached_prdctl, port_base + PORT_PRD_CTL); - /* issue r/w command */ - ap->ops->sff_exec_command(ap, &qc->tf); + readb(port_base + PORT_RPQ_FIFO); + readb(port_base + PORT_RPQ_CNT); + writew(0, port_base + PORT_IDMA_CTL); } -static void inic_bmdma_start(struct ata_queued_cmd *qc) +static void inic_host_err_intr(struct ata_port *ap, u8 irq_stat, u16 idma_stat) { - struct ata_port *ap = qc->ap; + struct ata_eh_info *ehi = &ap->link.eh_info; struct inic_port_priv *pp = ap->private_data; - void __iomem *port_base = inic_port_base(ap); + struct inic_cpb *cpb = &pp->pkt->cpb; + bool freeze = false; - /* start host DMA transaction */ - pp->cached_prdctl |= PRD_CTL_START; - writeb(pp->cached_prdctl, port_base + PORT_PRD_CTL); -} + ata_ehi_clear_desc(ehi); + ata_ehi_push_desc(ehi, "irq_stat=0x%x idma_stat=0x%x", + irq_stat, idma_stat); -static void inic_bmdma_stop(struct ata_queued_cmd *qc) -{ - struct ata_port *ap = qc->ap; - struct inic_port_priv *pp = ap->private_data; - void __iomem *port_base = inic_port_base(ap); + inic_stop_idma(ap); - /* stop DMA engine */ - writeb(pp->dfl_prdctl, port_base + PORT_PRD_CTL); -} + if (irq_stat & (PIRQ_OFFLINE | PIRQ_ONLINE)) { + ata_ehi_push_desc(ehi, "hotplug"); + ata_ehi_hotplugged(ehi); + freeze = true; + } -static u8 inic_bmdma_status(struct ata_port *ap) -{ - /* event is already verified by the interrupt handler */ - return ATA_DMA_INTR; + if (idma_stat & IDMA_STAT_PERR) { + ata_ehi_push_desc(ehi, "PCI error"); + freeze = true; + } + + if (idma_stat & IDMA_STAT_CPBERR) { + ata_ehi_push_desc(ehi, "CPB error"); + + if (cpb->resp_flags & CPB_RESP_IGNORED) { + __ata_ehi_push_desc(ehi, " ignored"); + ehi->err_mask |= AC_ERR_INVALID; + freeze = true; + } + + if (cpb->resp_flags & CPB_RESP_ATA_ERR) + ehi->err_mask |= AC_ERR_DEV; + + if (cpb->resp_flags & CPB_RESP_SPURIOUS) { + __ata_ehi_push_desc(ehi, " spurious-intr"); + ehi->err_mask |= AC_ERR_HSM; + freeze = true; + } + + if (cpb->resp_flags & + (CPB_RESP_UNDERFLOW | CPB_RESP_OVERFLOW)) { + __ata_ehi_push_desc(ehi, " data-over/underflow"); + ehi->err_mask |= AC_ERR_HSM; + freeze = true; + } + } + + if (freeze) + ata_port_freeze(ap); + else + ata_port_abort(ap); } static void inic_host_intr(struct ata_port *ap) { void __iomem *port_base = inic_port_base(ap); - struct ata_eh_info *ehi = &ap->link.eh_info; + struct ata_queued_cmd *qc = ata_qc_from_tag(ap, ap->link.active_tag); u8 irq_stat; + u16 idma_stat; - /* fetch and clear irq */ + /* read and clear IRQ status */ irq_stat = readb(port_base + PORT_IRQ_STAT); writeb(irq_stat, port_base + PORT_IRQ_STAT); + idma_stat = readw(port_base + PORT_IDMA_STAT); - if (likely(!(irq_stat & PIRQ_ERR))) { - struct ata_queued_cmd *qc = - ata_qc_from_tag(ap, ap->link.active_tag); + if (unlikely((irq_stat & PIRQ_ERR) || (idma_stat & IDMA_STAT_ERR))) + inic_host_err_intr(ap, irq_stat, idma_stat); - if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) { - ap->ops->sff_check_status(ap); /* clear ATA interrupt */ - return; - } + if (unlikely(!qc)) + goto spurious; - if (likely(ata_sff_host_intr(ap, qc))) - return; + if (likely(idma_stat & IDMA_STAT_DONE)) { + inic_stop_idma(ap); - ap->ops->sff_check_status(ap); /* clear ATA interrupt */ - ata_port_printk(ap, KERN_WARNING, "unhandled " - "interrupt, irq_stat=%x\n", irq_stat); + /* Depending on circumstances, device error + * isn't reported by IDMA, check it explicitly. + */ + if (unlikely(readb(port_base + PORT_TF_COMMAND) & + (ATA_DF | ATA_ERR))) + qc->err_mask |= AC_ERR_DEV; + + ata_qc_complete(qc); return; } - /* error */ - ata_ehi_push_desc(ehi, "irq_stat=0x%x", irq_stat); - - if (irq_stat & (PIRQ_OFFLINE | PIRQ_ONLINE)) { - ata_ehi_hotplugged(ehi); - ata_port_freeze(ap); - } else - ata_port_abort(ap); + spurious: + ata_port_printk(ap, KERN_WARNING, "unhandled interrupt: " + "cmd=0x%x irq_stat=0x%x idma_stat=0x%x\n", + qc ? qc->tf.command : 0xff, irq_stat, idma_stat); } static irqreturn_t inic_interrupt(int irq, void *dev_instance) { struct ata_host *host = dev_instance; - void __iomem *mmio_base = host->iomap[MMIO_BAR]; + struct inic_host_priv *hpriv = host->private_data; u16 host_irq_stat; int i, handled = 0;; - host_irq_stat = readw(mmio_base + HOST_IRQ_STAT); + host_irq_stat = readw(hpriv->mmio_base + HOST_IRQ_STAT); if (unlikely(!(host_irq_stat & HIRQ_GLOBAL))) goto out; @@ -327,60 +436,173 @@ static irqreturn_t inic_interrupt(int irq, void *dev_instance) return IRQ_RETVAL(handled); } +static int inic_check_atapi_dma(struct ata_queued_cmd *qc) +{ + /* For some reason ATAPI_PROT_DMA doesn't work for some + * commands including writes and other misc ops. Use PIO + * protocol instead, which BTW is driven by the DMA engine + * anyway, so it shouldn't make much difference for native + * SATA devices. + */ + if (atapi_cmd_type(qc->cdb[0]) == READ) + return 0; + return 1; +} + +static void inic_fill_sg(struct inic_prd *prd, struct ata_queued_cmd *qc) +{ + struct scatterlist *sg; + unsigned int si; + u8 flags = 0; + + if (qc->tf.flags & ATA_TFLAG_WRITE) + flags |= PRD_WRITE; + + if (ata_is_dma(qc->tf.protocol)) + flags |= PRD_DMA; + + for_each_sg(qc->sg, sg, qc->n_elem, si) { + prd->mad = cpu_to_le32(sg_dma_address(sg)); + prd->len = cpu_to_le16(sg_dma_len(sg)); + prd->flags = flags; + prd++; + } + + WARN_ON(!si); + prd[-1].flags |= PRD_END; +} + +static void inic_qc_prep(struct ata_queued_cmd *qc) +{ + struct inic_port_priv *pp = qc->ap->private_data; + struct inic_pkt *pkt = pp->pkt; + struct inic_cpb *cpb = &pkt->cpb; + struct inic_prd *prd = pkt->prd; + bool is_atapi = ata_is_atapi(qc->tf.protocol); + bool is_data = ata_is_data(qc->tf.protocol); + unsigned int cdb_len = 0; + + VPRINTK("ENTER\n"); + + if (is_atapi) + cdb_len = qc->dev->cdb_len; + + /* prepare packet, based on initio driver */ + memset(pkt, 0, sizeof(struct inic_pkt)); + + cpb->ctl_flags = CPB_CTL_VALID | CPB_CTL_IEN; + if (is_atapi || is_data) + cpb->ctl_flags |= CPB_CTL_DATA; + + cpb->len = cpu_to_le32(qc->nbytes + cdb_len); + cpb->prd = cpu_to_le32(pp->pkt_dma + offsetof(struct inic_pkt, prd)); + + cpb->device = qc->tf.device; + cpb->feature = qc->tf.feature; + cpb->nsect = qc->tf.nsect; + cpb->lbal = qc->tf.lbal; + cpb->lbam = qc->tf.lbam; + cpb->lbah = qc->tf.lbah; + + if (qc->tf.flags & ATA_TFLAG_LBA48) { + cpb->hob_feature = qc->tf.hob_feature; + cpb->hob_nsect = qc->tf.hob_nsect; + cpb->hob_lbal = qc->tf.hob_lbal; + cpb->hob_lbam = qc->tf.hob_lbam; + cpb->hob_lbah = qc->tf.hob_lbah; + } + + cpb->command = qc->tf.command; + /* don't load ctl - dunno why. it's like that in the initio driver */ + + /* setup PRD for CDB */ + if (is_atapi) { + memcpy(pkt->cdb, qc->cdb, ATAPI_CDB_LEN); + prd->mad = cpu_to_le32(pp->pkt_dma + + offsetof(struct inic_pkt, cdb)); + prd->len = cpu_to_le16(cdb_len); + prd->flags = PRD_CDB | PRD_WRITE; + if (!is_data) + prd->flags |= PRD_END; + prd++; + } + + /* setup sg table */ + if (is_data) + inic_fill_sg(prd, qc); + + pp->cpb_tbl[0] = pp->pkt_dma; +} + static unsigned int inic_qc_issue(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; + void __iomem *port_base = inic_port_base(ap); - /* ATA IRQ doesn't wait for DMA transfer completion and vice - * versa. Mask IRQ selectively to detect command completion. - * Without it, ATA DMA read command can cause data corruption. - * - * Something similar might be needed for ATAPI writes. I - * tried a lot of combinations but couldn't find the solution. - */ - if (qc->tf.protocol == ATA_PROT_DMA && - !(qc->tf.flags & ATA_TFLAG_WRITE)) - inic_set_pirq_mask(ap, PIRQ_MASK_DMA_READ); - else - inic_set_pirq_mask(ap, PIRQ_MASK_OTHER); + /* fire up the ADMA engine */ + writew(HCTL_FTHD0, port_base + HOST_CTL); + writew(IDMA_CTL_GO, port_base + PORT_IDMA_CTL); + writeb(0, port_base + PORT_CPB_PTQFIFO); + + return 0; +} + +static void inic_tf_read(struct ata_port *ap, struct ata_taskfile *tf) +{ + void __iomem *port_base = inic_port_base(ap); + + tf->feature = readb(port_base + PORT_TF_FEATURE); + tf->nsect = readb(port_base + PORT_TF_NSECT); + tf->lbal = readb(port_base + PORT_TF_LBAL); + tf->lbam = readb(port_base + PORT_TF_LBAM); + tf->lbah = readb(port_base + PORT_TF_LBAH); + tf->device = readb(port_base + PORT_TF_DEVICE); + tf->command = readb(port_base + PORT_TF_COMMAND); +} - /* Issuing a command to yet uninitialized port locks up the - * controller. Most of the time, this happens for the first - * command after reset which are ATA and ATAPI IDENTIFYs. - * Fast fail if stat is 0x7f or 0xff for those commands. +static bool inic_qc_fill_rtf(struct ata_queued_cmd *qc) +{ + struct ata_taskfile *rtf = &qc->result_tf; + struct ata_taskfile tf; + + /* FIXME: Except for status and error, result TF access + * doesn't work. I tried reading from BAR0/2, CPB and BAR5. + * None works regardless of which command interface is used. + * For now return true iff status indicates device error. + * This means that we're reporting bogus sector for RW + * failures. Eeekk.... */ - if (unlikely(qc->tf.command == ATA_CMD_ID_ATA || - qc->tf.command == ATA_CMD_ID_ATAPI)) { - u8 stat = ap->ops->sff_check_status(ap); - if (stat == 0x7f || stat == 0xff) - return AC_ERR_HSM; - } + inic_tf_read(qc->ap, &tf); - return ata_sff_qc_issue(qc); + if (!(tf.command & ATA_ERR)) + return false; + + rtf->command = tf.command; + rtf->feature = tf.feature; + return true; } static void inic_freeze(struct ata_port *ap) { void __iomem *port_base = inic_port_base(ap); - __inic_set_pirq_mask(ap, PIRQ_MASK_FREEZE); - - ap->ops->sff_check_status(ap); + writeb(PIRQ_MASK_FREEZE, port_base + PORT_IRQ_MASK); writeb(0xff, port_base + PORT_IRQ_STAT); - - readb(port_base + PORT_IRQ_STAT); /* flush */ } static void inic_thaw(struct ata_port *ap) { void __iomem *port_base = inic_port_base(ap); - ap->ops->sff_check_status(ap); writeb(0xff, port_base + PORT_IRQ_STAT); + writeb(PIRQ_MASK_DEFAULT, port_base + PORT_IRQ_MASK); +} - __inic_set_pirq_mask(ap, PIRQ_MASK_OTHER); +static int inic_check_ready(struct ata_link *link) +{ + void __iomem *port_base = inic_port_base(link->ap); - readb(port_base + PORT_IRQ_STAT); /* flush */ + return ata_check_ready(readb(port_base + PORT_TF_COMMAND)); } /* @@ -394,17 +616,15 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class, void __iomem *port_base = inic_port_base(ap); void __iomem *idma_ctl = port_base + PORT_IDMA_CTL; const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context); - u16 val; int rc; /* hammer it into sane state */ inic_reset_port(port_base); - val = readw(idma_ctl); - writew(val | IDMA_CTL_RST_ATA, idma_ctl); + writew(IDMA_CTL_RST_ATA, idma_ctl); readw(idma_ctl); /* flush */ msleep(1); - writew(val & ~IDMA_CTL_RST_ATA, idma_ctl); + writew(0, idma_ctl); rc = sata_link_resume(link, timing, deadline); if (rc) { @@ -418,7 +638,7 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class, struct ata_taskfile tf; /* wait for link to become ready */ - rc = ata_sff_wait_after_reset(link, 1, deadline); + rc = ata_wait_after_reset(link, deadline, inic_check_ready); /* link occupied, -ENODEV too is an error */ if (rc) { ata_link_printk(link, KERN_WARNING, "device not ready " @@ -426,7 +646,7 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class, return rc; } - ata_sff_tf_read(ap, &tf); + inic_tf_read(ap, &tf); *class = ata_dev_classify(&tf); } @@ -436,18 +656,8 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class, static void inic_error_handler(struct ata_port *ap) { void __iomem *port_base = inic_port_base(ap); - struct inic_port_priv *pp = ap->private_data; - unsigned long flags; - /* reset PIO HSM and stop DMA engine */ inic_reset_port(port_base); - - spin_lock_irqsave(ap->lock, flags); - ap->hsm_task_state = HSM_ST_IDLE; - writeb(pp->dfl_prdctl, port_base + PORT_PRD_CTL); - spin_unlock_irqrestore(ap->lock, flags); - - /* PIO and DMA engines have been stopped, perform recovery */ ata_std_error_handler(ap); } @@ -458,26 +668,18 @@ static void inic_post_internal_cmd(struct ata_queued_cmd *qc) inic_reset_port(inic_port_base(qc->ap)); } -static void inic_dev_config(struct ata_device *dev) -{ - /* inic can only handle upto LBA28 max sectors */ - if (dev->max_sectors > ATA_MAX_SECTORS) - dev->max_sectors = ATA_MAX_SECTORS; - - if (dev->n_sectors >= 1 << 28) { - ata_dev_printk(dev, KERN_ERR, - "ERROR: This driver doesn't support LBA48 yet and may cause\n" - " data corruption on such devices. Disabling.\n"); - ata_dev_disable(dev); - } -} - static void init_port(struct ata_port *ap) { void __iomem *port_base = inic_port_base(ap); + struct inic_port_priv *pp = ap->private_data; - /* Setup PRD address */ + /* clear packet and CPB table */ + memset(pp->pkt, 0, sizeof(struct inic_pkt)); + memset(pp->cpb_tbl, 0, IDMA_CPB_TBL_SIZE); + + /* setup PRD and CPB lookup table addresses */ writel(ap->prd_dma, port_base + PORT_PRD_ADDR); + writel(pp->cpb_tbl_dma, port_base + PORT_CPB_CPBLAR); } static int inic_port_resume(struct ata_port *ap) @@ -488,28 +690,30 @@ static int inic_port_resume(struct ata_port *ap) static int inic_port_start(struct ata_port *ap) { - void __iomem *port_base = inic_port_base(ap); + struct device *dev = ap->host->dev; struct inic_port_priv *pp; - u8 tmp; int rc; /* alloc and initialize private data */ - pp = devm_kzalloc(ap->host->dev, sizeof(*pp), GFP_KERNEL); + pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL); if (!pp) return -ENOMEM; ap->private_data = pp; - /* default PRD_CTL value, DMAEN, WR and START off */ - tmp = readb(port_base + PORT_PRD_CTL); - tmp &= ~(PRD_CTL_DMAEN | PRD_CTL_WR | PRD_CTL_START); - pp->dfl_prdctl = tmp; - /* Alloc resources */ rc = ata_port_start(ap); - if (rc) { - kfree(pp); + if (rc) return rc; - } + + pp->pkt = dmam_alloc_coherent(dev, sizeof(struct inic_pkt), + &pp->pkt_dma, GFP_KERNEL); + if (!pp->pkt) + return -ENOMEM; + + pp->cpb_tbl = dmam_alloc_coherent(dev, IDMA_CPB_TBL_SIZE, + &pp->cpb_tbl_dma, GFP_KERNEL); + if (!pp->cpb_tbl) + return -ENOMEM; init_port(ap); @@ -517,21 +721,18 @@ static int inic_port_start(struct ata_port *ap) } static struct ata_port_operations inic_port_ops = { - .inherits = &ata_sff_port_ops, + .inherits = &sata_port_ops, - .bmdma_setup = inic_bmdma_setup, - .bmdma_start = inic_bmdma_start, - .bmdma_stop = inic_bmdma_stop, - .bmdma_status = inic_bmdma_status, + .check_atapi_dma = inic_check_atapi_dma, + .qc_prep = inic_qc_prep, .qc_issue = inic_qc_issue, + .qc_fill_rtf = inic_qc_fill_rtf, .freeze = inic_freeze, .thaw = inic_thaw, - .softreset = ATA_OP_NULL, /* softreset is broken */ .hardreset = inic_hardreset, .error_handler = inic_error_handler, .post_internal_cmd = inic_post_internal_cmd, - .dev_config = inic_dev_config, .scr_read = inic_scr_read, .scr_write = inic_scr_write, @@ -541,12 +742,6 @@ static struct ata_port_operations inic_port_ops = { }; static struct ata_port_info inic_port_info = { - /* For some reason, ATAPI_PROT_PIO is broken on this - * controller, and no, PIO_POLLING does't fix it. It somehow - * manages to report the wrong ireason and ignoring ireason - * results in machine lock up. Tell libata to always prefer - * DMA. - */ .flags = ATA_FLAG_SATA | ATA_FLAG_PIO_DMA, .pio_mask = 0x1f, /* pio0-4 */ .mwdma_mask = 0x07, /* mwdma0-2 */ @@ -599,7 +794,6 @@ static int inic_pci_device_resume(struct pci_dev *pdev) { struct ata_host *host = dev_get_drvdata(&pdev->dev); struct inic_host_priv *hpriv = host->private_data; - void __iomem *mmio_base = host->iomap[MMIO_BAR]; int rc; rc = ata_pci_device_do_resume(pdev); @@ -607,7 +801,7 @@ static int inic_pci_device_resume(struct pci_dev *pdev) return rc; if (pdev->dev.power.power_state.event == PM_EVENT_SUSPEND) { - rc = init_controller(mmio_base, hpriv->cached_hctl); + rc = init_controller(hpriv->mmio_base, hpriv->cached_hctl); if (rc) return rc; } @@ -625,6 +819,7 @@ static int inic_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) struct ata_host *host; struct inic_host_priv *hpriv; void __iomem * const *iomap; + int mmio_bar; int i, rc; if (!printed_version++) @@ -638,38 +833,31 @@ static int inic_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) host->private_data = hpriv; - /* acquire resources and fill host */ + /* Acquire resources and fill host. Note that PCI and cardbus + * use different BARs. + */ rc = pcim_enable_device(pdev); if (rc) return rc; - rc = pcim_iomap_regions(pdev, 0x3f, DRV_NAME); + if (pci_resource_flags(pdev, MMIO_BAR_PCI) & IORESOURCE_MEM) + mmio_bar = MMIO_BAR_PCI; + else + mmio_bar = MMIO_BAR_CARDBUS; + + rc = pcim_iomap_regions(pdev, 1 << mmio_bar, DRV_NAME); if (rc) return rc; host->iomap = iomap = pcim_iomap_table(pdev); + hpriv->mmio_base = iomap[mmio_bar]; + hpriv->cached_hctl = readw(hpriv->mmio_base + HOST_CTL); for (i = 0; i < NR_PORTS; i++) { struct ata_port *ap = host->ports[i]; - struct ata_ioports *port = &ap->ioaddr; - unsigned int offset = i * PORT_SIZE; - - port->cmd_addr = iomap[2 * i]; - port->altstatus_addr = - port->ctl_addr = (void __iomem *) - ((unsigned long)iomap[2 * i + 1] | ATA_PCI_CTL_OFS); - port->scr_addr = iomap[MMIO_BAR] + offset + PORT_SCR; - - ata_sff_std_ports(port); - - ata_port_pbar_desc(ap, MMIO_BAR, -1, "mmio"); - ata_port_pbar_desc(ap, MMIO_BAR, offset, "port"); - ata_port_desc(ap, "cmd 0x%llx ctl 0x%llx", - (unsigned long long)pci_resource_start(pdev, 2 * i), - (unsigned long long)pci_resource_start(pdev, (2 * i + 1)) | - ATA_PCI_CTL_OFS); - } - hpriv->cached_hctl = readw(iomap[MMIO_BAR] + HOST_CTL); + ata_port_pbar_desc(ap, mmio_bar, -1, "mmio"); + ata_port_pbar_desc(ap, mmio_bar, i * PORT_SIZE, "port"); + } /* Set dma_mask. This devices doesn't support 64bit addressing. */ rc = pci_set_dma_mask(pdev, DMA_32BIT_MASK); @@ -698,7 +886,7 @@ static int inic_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) return rc; } - rc = init_controller(iomap[MMIO_BAR], hpriv->cached_hctl); + rc = init_controller(hpriv->mmio_base, hpriv->cached_hctl); if (rc) { dev_printk(KERN_ERR, &pdev->dev, "failed to initialize controller\n"); diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c index 842b1a15b78..bb73b222262 100644 --- a/drivers/ata/sata_mv.c +++ b/drivers/ata/sata_mv.c @@ -65,6 +65,7 @@ #include <linux/platform_device.h> #include <linux/ata_platform.h> #include <linux/mbus.h> +#include <linux/bitops.h> #include <scsi/scsi_host.h> #include <scsi/scsi_cmnd.h> #include <scsi/scsi_device.h> @@ -91,9 +92,9 @@ enum { MV_IRQ_COAL_TIME_THRESHOLD = (MV_IRQ_COAL_REG_BASE + 0xd0), MV_SATAHC0_REG_BASE = 0x20000, - MV_FLASH_CTL = 0x1046c, - MV_GPIO_PORT_CTL = 0x104f0, - MV_RESET_CFG = 0x180d8, + MV_FLASH_CTL_OFS = 0x1046c, + MV_GPIO_PORT_CTL_OFS = 0x104f0, + MV_RESET_CFG_OFS = 0x180d8, MV_PCI_REG_SZ = MV_MAJOR_REG_AREA_SZ, MV_SATAHC_REG_SZ = MV_MAJOR_REG_AREA_SZ, @@ -147,18 +148,21 @@ enum { /* PCI interface registers */ PCI_COMMAND_OFS = 0xc00, + PCI_COMMAND_MRDTRIG = (1 << 7), /* PCI Master Read Trigger */ PCI_MAIN_CMD_STS_OFS = 0xd30, STOP_PCI_MASTER = (1 << 2), PCI_MASTER_EMPTY = (1 << 3), GLOB_SFT_RST = (1 << 4), - MV_PCI_MODE = 0xd00, + MV_PCI_MODE_OFS = 0xd00, + MV_PCI_MODE_MASK = 0x30, + MV_PCI_EXP_ROM_BAR_CTL = 0xd2c, MV_PCI_DISC_TIMER = 0xd04, MV_PCI_MSI_TRIGGER = 0xc38, MV_PCI_SERR_MASK = 0xc28, - MV_PCI_XBAR_TMOUT = 0x1d04, + MV_PCI_XBAR_TMOUT_OFS = 0x1d04, MV_PCI_ERR_LOW_ADDRESS = 0x1d40, MV_PCI_ERR_HIGH_ADDRESS = 0x1d44, MV_PCI_ERR_ATTRIBUTE = 0x1d48, @@ -225,16 +229,18 @@ enum { PHY_MODE4 = 0x314, PHY_MODE2 = 0x330, SATA_IFCTL_OFS = 0x344, + SATA_TESTCTL_OFS = 0x348, SATA_IFSTAT_OFS = 0x34c, VENDOR_UNIQUE_FIS_OFS = 0x35c, - FIS_CFG_OFS = 0x360, - FIS_CFG_SINGLE_SYNC = (1 << 16), /* SYNC on DMA activation */ + FISCFG_OFS = 0x360, + FISCFG_WAIT_DEV_ERR = (1 << 8), /* wait for host on DevErr */ + FISCFG_SINGLE_SYNC = (1 << 16), /* SYNC on DMA activation */ MV5_PHY_MODE = 0x74, - MV5_LT_MODE = 0x30, - MV5_PHY_CTL = 0x0C, - SATA_INTERFACE_CFG = 0x050, + MV5_LTMODE_OFS = 0x30, + MV5_PHY_CTL_OFS = 0x0C, + SATA_INTERFACE_CFG_OFS = 0x050, MV_M2_PREAMP_MASK = 0x7e0, @@ -332,10 +338,16 @@ enum { EDMA_CMD_OFS = 0x28, /* EDMA command register */ EDMA_EN = (1 << 0), /* enable EDMA */ EDMA_DS = (1 << 1), /* disable EDMA; self-negated */ - ATA_RST = (1 << 2), /* reset trans/link/phy */ + EDMA_RESET = (1 << 2), /* reset eng/trans/link/phy */ + + EDMA_STATUS_OFS = 0x30, /* EDMA engine status */ + EDMA_STATUS_CACHE_EMPTY = (1 << 6), /* GenIIe command cache empty */ + EDMA_STATUS_IDLE = (1 << 7), /* GenIIe EDMA enabled/idle */ - EDMA_IORDY_TMOUT = 0x34, - EDMA_ARB_CFG = 0x38, + EDMA_IORDY_TMOUT_OFS = 0x34, + EDMA_ARB_CFG_OFS = 0x38, + + EDMA_HALTCOND_OFS = 0x60, /* GenIIe halt conditions */ GEN_II_NCQ_MAX_SECTORS = 256, /* max sects/io on Gen2 w/NCQ */ @@ -350,15 +362,19 @@ enum { MV_HP_GEN_II = (1 << 7), /* Generation II: 60xx */ MV_HP_GEN_IIE = (1 << 8), /* Generation IIE: 6042/7042 */ MV_HP_PCIE = (1 << 9), /* PCIe bus/regs: 7042 */ + MV_HP_CUT_THROUGH = (1 << 10), /* can use EDMA cut-through */ /* Port private flags (pp_flags) */ MV_PP_FLAG_EDMA_EN = (1 << 0), /* is EDMA engine enabled? */ MV_PP_FLAG_NCQ_EN = (1 << 1), /* is EDMA set up for NCQ? */ + MV_PP_FLAG_FBS_EN = (1 << 2), /* is EDMA set up for FBS? */ + MV_PP_FLAG_DELAYED_EH = (1 << 3), /* delayed dev err handling */ }; #define IS_GEN_I(hpriv) ((hpriv)->hp_flags & MV_HP_GEN_I) #define IS_GEN_II(hpriv) ((hpriv)->hp_flags & MV_HP_GEN_II) #define IS_GEN_IIE(hpriv) ((hpriv)->hp_flags & MV_HP_GEN_IIE) +#define IS_PCIE(hpriv) ((hpriv)->hp_flags & MV_HP_PCIE) #define HAS_PCI(host) (!((host)->ports[0]->flags & MV_FLAG_SOC)) #define WINDOW_CTRL(i) (0x20030 + ((i) << 4)) @@ -433,6 +449,7 @@ struct mv_port_priv { unsigned int resp_idx; u32 pp_flags; + unsigned int delayed_eh_pmp_map; }; struct mv_port_signal { @@ -479,6 +496,7 @@ static int mv5_scr_read(struct ata_port *ap, unsigned int sc_reg_in, u32 *val); static int mv5_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val); static int mv_port_start(struct ata_port *ap); static void mv_port_stop(struct ata_port *ap); +static int mv_qc_defer(struct ata_queued_cmd *qc); static void mv_qc_prep(struct ata_queued_cmd *qc); static void mv_qc_prep_iie(struct ata_queued_cmd *qc); static unsigned int mv_qc_issue(struct ata_queued_cmd *qc); @@ -527,6 +545,9 @@ static int mv_pmp_hardreset(struct ata_link *link, unsigned int *class, unsigned long deadline); static int mv_softreset(struct ata_link *link, unsigned int *class, unsigned long deadline); +static void mv_pmp_error_handler(struct ata_port *ap); +static void mv_process_crpb_entries(struct ata_port *ap, + struct mv_port_priv *pp); /* .sg_tablesize is (MV_MAX_SG_CT / 2) in the structures below * because we have to allow room for worst case splitting of @@ -548,6 +569,7 @@ static struct scsi_host_template mv6_sht = { static struct ata_port_operations mv5_ops = { .inherits = &ata_sff_port_ops, + .qc_defer = mv_qc_defer, .qc_prep = mv_qc_prep, .qc_issue = mv_qc_issue, @@ -566,7 +588,6 @@ static struct ata_port_operations mv5_ops = { static struct ata_port_operations mv6_ops = { .inherits = &mv5_ops, - .qc_defer = sata_pmp_qc_defer_cmd_switch, .dev_config = mv6_dev_config, .scr_read = mv_scr_read, .scr_write = mv_scr_write, @@ -574,12 +595,11 @@ static struct ata_port_operations mv6_ops = { .pmp_hardreset = mv_pmp_hardreset, .pmp_softreset = mv_softreset, .softreset = mv_softreset, - .error_handler = sata_pmp_error_handler, + .error_handler = mv_pmp_error_handler, }; static struct ata_port_operations mv_iie_ops = { .inherits = &mv6_ops, - .qc_defer = ata_std_qc_defer, /* FIS-based switching */ .dev_config = ATA_OP_NULL, .qc_prep = mv_qc_prep_iie, }; @@ -875,6 +895,29 @@ static void mv_start_dma(struct ata_port *ap, void __iomem *port_mmio, } } +static void mv_wait_for_edma_empty_idle(struct ata_port *ap) +{ + void __iomem *port_mmio = mv_ap_base(ap); + const u32 empty_idle = (EDMA_STATUS_CACHE_EMPTY | EDMA_STATUS_IDLE); + const int per_loop = 5, timeout = (15 * 1000 / per_loop); + int i; + + /* + * Wait for the EDMA engine to finish transactions in progress. + * No idea what a good "timeout" value might be, but measurements + * indicate that it often requires hundreds of microseconds + * with two drives in-use. So we use the 15msec value above + * as a rough guess at what even more drives might require. + */ + for (i = 0; i < timeout; ++i) { + u32 edma_stat = readl(port_mmio + EDMA_STATUS_OFS); + if ((edma_stat & empty_idle) == empty_idle) + break; + udelay(per_loop); + } + /* ata_port_printk(ap, KERN_INFO, "%s: %u+ usecs\n", __func__, i); */ +} + /** * mv_stop_edma_engine - Disable eDMA engine * @port_mmio: io base address @@ -907,6 +950,7 @@ static int mv_stop_edma(struct ata_port *ap) if (!(pp->pp_flags & MV_PP_FLAG_EDMA_EN)) return 0; pp->pp_flags &= ~MV_PP_FLAG_EDMA_EN; + mv_wait_for_edma_empty_idle(ap); if (mv_stop_edma_engine(port_mmio)) { ata_port_printk(ap, KERN_ERR, "Unable to stop eDMA\n"); return -EIO; @@ -1057,26 +1101,95 @@ static void mv6_dev_config(struct ata_device *adev) } } -static void mv_config_fbs(void __iomem *port_mmio, int enable_fbs) +static int mv_qc_defer(struct ata_queued_cmd *qc) { - u32 old_fcfg, new_fcfg, old_ltmode, new_ltmode; + struct ata_link *link = qc->dev->link; + struct ata_port *ap = link->ap; + struct mv_port_priv *pp = ap->private_data; + + /* + * Don't allow new commands if we're in a delayed EH state + * for NCQ and/or FIS-based switching. + */ + if (pp->pp_flags & MV_PP_FLAG_DELAYED_EH) + return ATA_DEFER_PORT; /* - * Various bit settings required for operation - * in FIS-based switching (fbs) mode on GenIIe: + * If the port is completely idle, then allow the new qc. */ - old_fcfg = readl(port_mmio + FIS_CFG_OFS); - old_ltmode = readl(port_mmio + LTMODE_OFS); - if (enable_fbs) { - new_fcfg = old_fcfg | FIS_CFG_SINGLE_SYNC; - new_ltmode = old_ltmode | LTMODE_BIT8; - } else { /* disable fbs */ - new_fcfg = old_fcfg & ~FIS_CFG_SINGLE_SYNC; - new_ltmode = old_ltmode & ~LTMODE_BIT8; - } - if (new_fcfg != old_fcfg) - writelfl(new_fcfg, port_mmio + FIS_CFG_OFS); + if (ap->nr_active_links == 0) + return 0; + + if (pp->pp_flags & MV_PP_FLAG_EDMA_EN) { + /* + * The port is operating in host queuing mode (EDMA). + * It can accomodate a new qc if the qc protocol + * is compatible with the current host queue mode. + */ + if (pp->pp_flags & MV_PP_FLAG_NCQ_EN) { + /* + * The host queue (EDMA) is in NCQ mode. + * If the new qc is also an NCQ command, + * then allow the new qc. + */ + if (qc->tf.protocol == ATA_PROT_NCQ) + return 0; + } else { + /* + * The host queue (EDMA) is in non-NCQ, DMA mode. + * If the new qc is also a non-NCQ, DMA command, + * then allow the new qc. + */ + if (qc->tf.protocol == ATA_PROT_DMA) + return 0; + } + } + return ATA_DEFER_PORT; +} + +static void mv_config_fbs(void __iomem *port_mmio, int want_ncq, int want_fbs) +{ + u32 new_fiscfg, old_fiscfg; + u32 new_ltmode, old_ltmode; + u32 new_haltcond, old_haltcond; + + old_fiscfg = readl(port_mmio + FISCFG_OFS); + old_ltmode = readl(port_mmio + LTMODE_OFS); + old_haltcond = readl(port_mmio + EDMA_HALTCOND_OFS); + + new_fiscfg = old_fiscfg & ~(FISCFG_SINGLE_SYNC | FISCFG_WAIT_DEV_ERR); + new_ltmode = old_ltmode & ~LTMODE_BIT8; + new_haltcond = old_haltcond | EDMA_ERR_DEV; + + if (want_fbs) { + new_fiscfg = old_fiscfg | FISCFG_SINGLE_SYNC; + new_ltmode = old_ltmode | LTMODE_BIT8; + if (want_ncq) + new_haltcond &= ~EDMA_ERR_DEV; + else + new_fiscfg |= FISCFG_WAIT_DEV_ERR; + } + + if (new_fiscfg != old_fiscfg) + writelfl(new_fiscfg, port_mmio + FISCFG_OFS); if (new_ltmode != old_ltmode) writelfl(new_ltmode, port_mmio + LTMODE_OFS); + if (new_haltcond != old_haltcond) + writelfl(new_haltcond, port_mmio + EDMA_HALTCOND_OFS); +} + +static void mv_60x1_errata_sata25(struct ata_port *ap, int want_ncq) +{ + struct mv_host_priv *hpriv = ap->host->private_data; + u32 old, new; + + /* workaround for 88SX60x1 FEr SATA#25 (part 1) */ + old = readl(hpriv->base + MV_GPIO_PORT_CTL_OFS); + if (want_ncq) + new = old | (1 << 22); + else + new = old & ~(1 << 22); + if (new != old) + writel(new, hpriv->base + MV_GPIO_PORT_CTL_OFS); } static void mv_edma_cfg(struct ata_port *ap, int want_ncq) @@ -1088,25 +1201,40 @@ static void mv_edma_cfg(struct ata_port *ap, int want_ncq) /* set up non-NCQ EDMA configuration */ cfg = EDMA_CFG_Q_DEPTH; /* always 0x1f for *all* chips */ + pp->pp_flags &= ~MV_PP_FLAG_FBS_EN; if (IS_GEN_I(hpriv)) cfg |= (1 << 8); /* enab config burst size mask */ - else if (IS_GEN_II(hpriv)) + else if (IS_GEN_II(hpriv)) { cfg |= EDMA_CFG_RD_BRST_EXT | EDMA_CFG_WR_BUFF_LEN; + mv_60x1_errata_sata25(ap, want_ncq); - else if (IS_GEN_IIE(hpriv)) { - cfg |= (1 << 23); /* do not mask PM field in rx'd FIS */ - cfg |= (1 << 22); /* enab 4-entry host queue cache */ - cfg |= (1 << 18); /* enab early completion */ - cfg |= (1 << 17); /* enab cut-through (dis stor&forwrd) */ + } else if (IS_GEN_IIE(hpriv)) { + int want_fbs = sata_pmp_attached(ap); + /* + * Possible future enhancement: + * + * The chip can use FBS with non-NCQ, if we allow it, + * But first we need to have the error handling in place + * for this mode (datasheet section 7.3.15.4.2.3). + * So disallow non-NCQ FBS for now. + */ + want_fbs &= want_ncq; + + mv_config_fbs(port_mmio, want_ncq, want_fbs); - if (want_ncq && sata_pmp_attached(ap)) { + if (want_fbs) { + pp->pp_flags |= MV_PP_FLAG_FBS_EN; cfg |= EDMA_CFG_EDMA_FBS; /* FIS-based switching */ - mv_config_fbs(port_mmio, 1); - } else { - mv_config_fbs(port_mmio, 0); } + + cfg |= (1 << 23); /* do not mask PM field in rx'd FIS */ + cfg |= (1 << 22); /* enab 4-entry host queue cache */ + if (HAS_PCI(ap->host)) + cfg |= (1 << 18); /* enab early completion */ + if (hpriv->hp_flags & MV_HP_CUT_THROUGH) + cfg |= (1 << 17); /* enab cut-thru (dis stor&forwrd) */ } if (want_ncq) { @@ -1483,25 +1611,186 @@ static struct ata_queued_cmd *mv_get_active_qc(struct ata_port *ap) return qc; } -static void mv_unexpected_intr(struct ata_port *ap) +static void mv_pmp_error_handler(struct ata_port *ap) { + unsigned int pmp, pmp_map; struct mv_port_priv *pp = ap->private_data; - struct ata_eh_info *ehi = &ap->link.eh_info; - char *when = ""; + if (pp->pp_flags & MV_PP_FLAG_DELAYED_EH) { + /* + * Perform NCQ error analysis on failed PMPs + * before we freeze the port entirely. + * + * The failed PMPs are marked earlier by mv_pmp_eh_prep(). + */ + pmp_map = pp->delayed_eh_pmp_map; + pp->pp_flags &= ~MV_PP_FLAG_DELAYED_EH; + for (pmp = 0; pmp_map != 0; pmp++) { + unsigned int this_pmp = (1 << pmp); + if (pmp_map & this_pmp) { + struct ata_link *link = &ap->pmp_link[pmp]; + pmp_map &= ~this_pmp; + ata_eh_analyze_ncq_error(link); + } + } + ata_port_freeze(ap); + } + sata_pmp_error_handler(ap); +} + +static unsigned int mv_get_err_pmp_map(struct ata_port *ap) +{ + void __iomem *port_mmio = mv_ap_base(ap); + + return readl(port_mmio + SATA_TESTCTL_OFS) >> 16; +} + +static void mv_pmp_eh_prep(struct ata_port *ap, unsigned int pmp_map) +{ + struct ata_eh_info *ehi; + unsigned int pmp; + + /* + * Initialize EH info for PMPs which saw device errors + */ + ehi = &ap->link.eh_info; + for (pmp = 0; pmp_map != 0; pmp++) { + unsigned int this_pmp = (1 << pmp); + if (pmp_map & this_pmp) { + struct ata_link *link = &ap->pmp_link[pmp]; + + pmp_map &= ~this_pmp; + ehi = &link->eh_info; + ata_ehi_clear_desc(ehi); + ata_ehi_push_desc(ehi, "dev err"); + ehi->err_mask |= AC_ERR_DEV; + ehi->action |= ATA_EH_RESET; + ata_link_abort(link); + } + } +} + +static int mv_handle_fbs_ncq_dev_err(struct ata_port *ap) +{ + struct mv_port_priv *pp = ap->private_data; + int failed_links; + unsigned int old_map, new_map; + + /* + * Device error during FBS+NCQ operation: + * + * Set a port flag to prevent further I/O being enqueued. + * Leave the EDMA running to drain outstanding commands from this port. + * Perform the post-mortem/EH only when all responses are complete. + * Follow recovery sequence from 6042/7042 datasheet (7.3.15.4.2.2). + */ + if (!(pp->pp_flags & MV_PP_FLAG_DELAYED_EH)) { + pp->pp_flags |= MV_PP_FLAG_DELAYED_EH; + pp->delayed_eh_pmp_map = 0; + } + old_map = pp->delayed_eh_pmp_map; + new_map = old_map | mv_get_err_pmp_map(ap); + + if (old_map != new_map) { + pp->delayed_eh_pmp_map = new_map; + mv_pmp_eh_prep(ap, new_map & ~old_map); + } + failed_links = hweight16(new_map); + + ata_port_printk(ap, KERN_INFO, "%s: pmp_map=%04x qc_map=%04x " + "failed_links=%d nr_active_links=%d\n", + __func__, pp->delayed_eh_pmp_map, + ap->qc_active, failed_links, + ap->nr_active_links); + + if (ap->nr_active_links <= failed_links) { + mv_process_crpb_entries(ap, pp); + mv_stop_edma(ap); + mv_eh_freeze(ap); + ata_port_printk(ap, KERN_INFO, "%s: done\n", __func__); + return 1; /* handled */ + } + ata_port_printk(ap, KERN_INFO, "%s: waiting\n", __func__); + return 1; /* handled */ +} + +static int mv_handle_fbs_non_ncq_dev_err(struct ata_port *ap) +{ /* - * We got a device interrupt from something that - * was supposed to be using EDMA or polling. + * Possible future enhancement: + * + * FBS+non-NCQ operation is not yet implemented. + * See related notes in mv_edma_cfg(). + * + * Device error during FBS+non-NCQ operation: + * + * We need to snapshot the shadow registers for each failed command. + * Follow recovery sequence from 6042/7042 datasheet (7.3.15.4.2.3). */ + return 0; /* not handled */ +} + +static int mv_handle_dev_err(struct ata_port *ap, u32 edma_err_cause) +{ + struct mv_port_priv *pp = ap->private_data; + + if (!(pp->pp_flags & MV_PP_FLAG_EDMA_EN)) + return 0; /* EDMA was not active: not handled */ + if (!(pp->pp_flags & MV_PP_FLAG_FBS_EN)) + return 0; /* FBS was not active: not handled */ + + if (!(edma_err_cause & EDMA_ERR_DEV)) + return 0; /* non DEV error: not handled */ + edma_err_cause &= ~EDMA_ERR_IRQ_TRANSIENT; + if (edma_err_cause & ~(EDMA_ERR_DEV | EDMA_ERR_SELF_DIS)) + return 0; /* other problems: not handled */ + + if (pp->pp_flags & MV_PP_FLAG_NCQ_EN) { + /* + * EDMA should NOT have self-disabled for this case. + * If it did, then something is wrong elsewhere, + * and we cannot handle it here. + */ + if (edma_err_cause & EDMA_ERR_SELF_DIS) { + ata_port_printk(ap, KERN_WARNING, + "%s: err_cause=0x%x pp_flags=0x%x\n", + __func__, edma_err_cause, pp->pp_flags); + return 0; /* not handled */ + } + return mv_handle_fbs_ncq_dev_err(ap); + } else { + /* + * EDMA should have self-disabled for this case. + * If it did not, then something is wrong elsewhere, + * and we cannot handle it here. + */ + if (!(edma_err_cause & EDMA_ERR_SELF_DIS)) { + ata_port_printk(ap, KERN_WARNING, + "%s: err_cause=0x%x pp_flags=0x%x\n", + __func__, edma_err_cause, pp->pp_flags); + return 0; /* not handled */ + } + return mv_handle_fbs_non_ncq_dev_err(ap); + } + return 0; /* not handled */ +} + +static void mv_unexpected_intr(struct ata_port *ap, int edma_was_enabled) +{ + struct ata_eh_info *ehi = &ap->link.eh_info; + char *when = "idle"; + ata_ehi_clear_desc(ehi); - if (pp->pp_flags & MV_PP_FLAG_EDMA_EN) { - when = " while EDMA enabled"; + if (!ap || (ap->flags & ATA_FLAG_DISABLED)) { + when = "disabled"; + } else if (edma_was_enabled) { + when = "EDMA enabled"; } else { struct ata_queued_cmd *qc = ata_qc_from_tag(ap, ap->link.active_tag); if (qc && (qc->tf.flags & ATA_TFLAG_POLLING)) - when = " while polling"; + when = "polling"; } - ata_ehi_push_desc(ehi, "unexpected device interrupt%s", when); + ata_ehi_push_desc(ehi, "unexpected device interrupt while %s", when); ehi->err_mask |= AC_ERR_OTHER; ehi->action |= ATA_EH_RESET; ata_port_freeze(ap); @@ -1519,7 +1808,7 @@ static void mv_unexpected_intr(struct ata_port *ap) * LOCKING: * Inherited from caller. */ -static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc) +static void mv_err_intr(struct ata_port *ap) { void __iomem *port_mmio = mv_ap_base(ap); u32 edma_err_cause, eh_freeze_mask, serr = 0; @@ -1527,24 +1816,42 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc) struct mv_host_priv *hpriv = ap->host->private_data; unsigned int action = 0, err_mask = 0; struct ata_eh_info *ehi = &ap->link.eh_info; - - ata_ehi_clear_desc(ehi); + struct ata_queued_cmd *qc; + int abort = 0; /* - * Read and clear the err_cause bits. This won't actually - * clear for some errors (eg. SError), but we will be doing - * a hard reset in those cases regardless, which *will* clear it. + * Read and clear the SError and err_cause bits. */ + sata_scr_read(&ap->link, SCR_ERROR, &serr); + sata_scr_write_flush(&ap->link, SCR_ERROR, serr); + edma_err_cause = readl(port_mmio + EDMA_ERR_IRQ_CAUSE_OFS); writelfl(~edma_err_cause, port_mmio + EDMA_ERR_IRQ_CAUSE_OFS); - ata_ehi_push_desc(ehi, "edma_err_cause=%08x", edma_err_cause); + ata_port_printk(ap, KERN_INFO, "%s: err_cause=%08x pp_flags=0x%x\n", + __func__, edma_err_cause, pp->pp_flags); + + if (edma_err_cause & EDMA_ERR_DEV) { + /* + * Device errors during FIS-based switching operation + * require special handling. + */ + if (mv_handle_dev_err(ap, edma_err_cause)) + return; + } + qc = mv_get_active_qc(ap); + ata_ehi_clear_desc(ehi); + ata_ehi_push_desc(ehi, "edma_err_cause=%08x pp_flags=%08x", + edma_err_cause, pp->pp_flags); /* * All generations share these EDMA error cause bits: */ - if (edma_err_cause & EDMA_ERR_DEV) + if (edma_err_cause & EDMA_ERR_DEV) { err_mask |= AC_ERR_DEV; + action |= ATA_EH_RESET; + ata_ehi_push_desc(ehi, "dev error"); + } if (edma_err_cause & (EDMA_ERR_D_PAR | EDMA_ERR_PRD_PAR | EDMA_ERR_CRQB_PAR | EDMA_ERR_CRPB_PAR | EDMA_ERR_INTRL_PAR)) { @@ -1576,13 +1883,6 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc) ata_ehi_push_desc(ehi, "EDMA self-disable"); } if (edma_err_cause & EDMA_ERR_SERR) { - /* - * Ensure that we read our own SCR, not a pmp link SCR: - */ - ap->ops->scr_read(ap, SCR_ERROR, &serr); - /* - * Don't clear SError here; leave it for libata-eh: - */ ata_ehi_push_desc(ehi, "SError=%08x", serr); err_mask |= AC_ERR_ATA_BUS; action |= ATA_EH_RESET; @@ -1602,10 +1902,29 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc) else ehi->err_mask |= err_mask; - if (edma_err_cause & eh_freeze_mask) + if (err_mask == AC_ERR_DEV) { + /* + * Cannot do ata_port_freeze() here, + * because it would kill PIO access, + * which is needed for further diagnosis. + */ + mv_eh_freeze(ap); + abort = 1; + } else if (edma_err_cause & eh_freeze_mask) { + /* + * Note to self: ata_port_freeze() calls ata_port_abort() + */ ata_port_freeze(ap); - else - ata_port_abort(ap); + } else { + abort = 1; + } + + if (abort) { + if (qc) + ata_link_abort(qc->dev->link); + else + ata_port_abort(ap); + } } static void mv_process_crpb_response(struct ata_port *ap, @@ -1632,8 +1951,9 @@ static void mv_process_crpb_response(struct ata_port *ap, } } ata_status = edma_status >> CRPB_FLAG_STATUS_SHIFT; - qc->err_mask |= ac_err_mask(ata_status); - ata_qc_complete(qc); + if (!ac_err_mask(ata_status)) + ata_qc_complete(qc); + /* else: leave it for mv_err_intr() */ } else { ata_port_printk(ap, KERN_ERR, "%s: no qc for tag=%d\n", __func__, tag); @@ -1677,6 +1997,44 @@ static void mv_process_crpb_entries(struct ata_port *ap, struct mv_port_priv *pp port_mmio + EDMA_RSP_Q_OUT_PTR_OFS); } +static void mv_port_intr(struct ata_port *ap, u32 port_cause) +{ + struct mv_port_priv *pp; + int edma_was_enabled; + + if (!ap || (ap->flags & ATA_FLAG_DISABLED)) { + mv_unexpected_intr(ap, 0); + return; + } + /* + * Grab a snapshot of the EDMA_EN flag setting, + * so that we have a consistent view for this port, + * even if something we call of our routines changes it. + */ + pp = ap->private_data; + edma_was_enabled = (pp->pp_flags & MV_PP_FLAG_EDMA_EN); + /* + * Process completed CRPB response(s) before other events. + */ + if (edma_was_enabled && (port_cause & DONE_IRQ)) { + mv_process_crpb_entries(ap, pp); + if (pp->pp_flags & MV_PP_FLAG_DELAYED_EH) + mv_handle_fbs_ncq_dev_err(ap); + } + /* + * Handle chip-reported errors, or continue on to handle PIO. + */ + if (unlikely(port_cause & ERR_IRQ)) { + mv_err_intr(ap); + } else if (!edma_was_enabled) { + struct ata_queued_cmd *qc = mv_get_active_qc(ap); + if (qc) + ata_sff_host_intr(ap, qc); + else + mv_unexpected_intr(ap, edma_was_enabled); + } +} + /** * mv_host_intr - Handle all interrupts on the given host controller * @host: host specific structure @@ -1688,66 +2046,58 @@ static void mv_process_crpb_entries(struct ata_port *ap, struct mv_port_priv *pp static int mv_host_intr(struct ata_host *host, u32 main_irq_cause) { struct mv_host_priv *hpriv = host->private_data; - void __iomem *mmio = hpriv->base, *hc_mmio = NULL; - u32 hc_irq_cause = 0; + void __iomem *mmio = hpriv->base, *hc_mmio; unsigned int handled = 0, port; for (port = 0; port < hpriv->n_ports; port++) { struct ata_port *ap = host->ports[port]; - struct mv_port_priv *pp; - unsigned int shift, hardport, port_cause; - /* - * When we move to the second hc, flag our cached - * copies of hc_mmio (and hc_irq_cause) as invalid again. - */ - if (port == MV_PORTS_PER_HC) - hc_mmio = NULL; - /* - * Do nothing if port is not interrupting or is disabled: - */ + unsigned int p, shift, hardport, port_cause; + MV_PORT_TO_SHIFT_AND_HARDPORT(port, shift, hardport); - port_cause = (main_irq_cause >> shift) & (DONE_IRQ | ERR_IRQ); - if (!port_cause || !ap || (ap->flags & ATA_FLAG_DISABLED)) - continue; /* - * Each hc within the host has its own hc_irq_cause register. - * We defer reading it until we know we need it, right now: - * - * FIXME later: we don't really need to read this register - * (some logic changes required below if we go that way), - * because it doesn't tell us anything new. But we do need - * to write to it, outside the top of this loop, - * to reset the interrupt triggers for next time. + * Each hc within the host has its own hc_irq_cause register, + * where the interrupting ports bits get ack'd. */ - if (!hc_mmio) { + if (hardport == 0) { /* first port on this hc ? */ + u32 hc_cause = (main_irq_cause >> shift) & HC0_IRQ_PEND; + u32 port_mask, ack_irqs; + /* + * Skip this entire hc if nothing pending for any ports + */ + if (!hc_cause) { + port += MV_PORTS_PER_HC - 1; + continue; + } + /* + * We don't need/want to read the hc_irq_cause register, + * because doing so hurts performance, and + * main_irq_cause already gives us everything we need. + * + * But we do have to *write* to the hc_irq_cause to ack + * the ports that we are handling this time through. + * + * This requires that we create a bitmap for those + * ports which interrupted us, and use that bitmap + * to ack (only) those ports via hc_irq_cause. + */ + ack_irqs = 0; + for (p = 0; p < MV_PORTS_PER_HC; ++p) { + if ((port + p) >= hpriv->n_ports) + break; + port_mask = (DONE_IRQ | ERR_IRQ) << (p * 2); + if (hc_cause & port_mask) + ack_irqs |= (DMA_IRQ | DEV_IRQ) << p; + } hc_mmio = mv_hc_base_from_port(mmio, port); - hc_irq_cause = readl(hc_mmio + HC_IRQ_CAUSE_OFS); - writelfl(~hc_irq_cause, hc_mmio + HC_IRQ_CAUSE_OFS); + writelfl(~ack_irqs, hc_mmio + HC_IRQ_CAUSE_OFS); handled = 1; } /* - * Process completed CRPB response(s) before other events. - */ - pp = ap->private_data; - if (hc_irq_cause & (DMA_IRQ << hardport)) { - if (pp->pp_flags & MV_PP_FLAG_EDMA_EN) - mv_process_crpb_entries(ap, pp); - } - /* - * Handle chip-reported errors, or continue on to handle PIO. + * Handle interrupts signalled for this port: */ - if (unlikely(port_cause & ERR_IRQ)) { - mv_err_intr(ap, mv_get_active_qc(ap)); - } else if (hc_irq_cause & (DEV_IRQ << hardport)) { - if (!(pp->pp_flags & MV_PP_FLAG_EDMA_EN)) { - struct ata_queued_cmd *qc = mv_get_active_qc(ap); - if (qc) { - ata_sff_host_intr(ap, qc); - continue; - } - } - mv_unexpected_intr(ap); - } + port_cause = (main_irq_cause >> shift) & (DONE_IRQ | ERR_IRQ); + if (port_cause) + mv_port_intr(ap, port_cause); } return handled; } @@ -1894,7 +2244,7 @@ static void mv5_reset_bus(struct ata_host *host, void __iomem *mmio) static void mv5_reset_flash(struct mv_host_priv *hpriv, void __iomem *mmio) { - writel(0x0fcfffff, mmio + MV_FLASH_CTL); + writel(0x0fcfffff, mmio + MV_FLASH_CTL_OFS); } static void mv5_read_preamp(struct mv_host_priv *hpriv, int idx, @@ -1913,7 +2263,7 @@ static void mv5_enable_leds(struct mv_host_priv *hpriv, void __iomem *mmio) { u32 tmp; - writel(0, mmio + MV_GPIO_PORT_CTL); + writel(0, mmio + MV_GPIO_PORT_CTL_OFS); /* FIXME: handle MV_HP_ERRATA_50XXB2 errata */ @@ -1931,14 +2281,14 @@ static void mv5_phy_errata(struct mv_host_priv *hpriv, void __iomem *mmio, int fix_apm_sq = (hpriv->hp_flags & MV_HP_ERRATA_50XXB0); if (fix_apm_sq) { - tmp = readl(phy_mmio + MV5_LT_MODE); + tmp = readl(phy_mmio + MV5_LTMODE_OFS); tmp |= (1 << 19); - writel(tmp, phy_mmio + MV5_LT_MODE); + writel(tmp, phy_mmio + MV5_LTMODE_OFS); - tmp = readl(phy_mmio + MV5_PHY_CTL); + tmp = readl(phy_mmio + MV5_PHY_CTL_OFS); tmp &= ~0x3; tmp |= 0x1; - writel(tmp, phy_mmio + MV5_PHY_CTL); + writel(tmp, phy_mmio + MV5_PHY_CTL_OFS); } tmp = readl(phy_mmio + MV5_PHY_MODE); @@ -1956,11 +2306,6 @@ static void mv5_reset_hc_port(struct mv_host_priv *hpriv, void __iomem *mmio, { void __iomem *port_mmio = mv_port_base(mmio, port); - /* - * The datasheet warns against setting ATA_RST when EDMA is active - * (but doesn't say what the problem might be). So we first try - * to disable the EDMA engine before doing the ATA_RST operation. - */ mv_reset_channel(hpriv, mmio, port); ZERO(0x028); /* command */ @@ -1975,7 +2320,7 @@ static void mv5_reset_hc_port(struct mv_host_priv *hpriv, void __iomem *mmio, ZERO(0x024); /* respq outp */ ZERO(0x020); /* respq inp */ ZERO(0x02c); /* test control */ - writel(0xbc, port_mmio + EDMA_IORDY_TMOUT); + writel(0xbc, port_mmio + EDMA_IORDY_TMOUT_OFS); } #undef ZERO @@ -2021,13 +2366,13 @@ static void mv_reset_pci_bus(struct ata_host *host, void __iomem *mmio) struct mv_host_priv *hpriv = host->private_data; u32 tmp; - tmp = readl(mmio + MV_PCI_MODE); + tmp = readl(mmio + MV_PCI_MODE_OFS); tmp &= 0xff00ffff; - writel(tmp, mmio + MV_PCI_MODE); + writel(tmp, mmio + MV_PCI_MODE_OFS); ZERO(MV_PCI_DISC_TIMER); ZERO(MV_PCI_MSI_TRIGGER); - writel(0x000100ff, mmio + MV_PCI_XBAR_TMOUT); + writel(0x000100ff, mmio + MV_PCI_XBAR_TMOUT_OFS); ZERO(PCI_HC_MAIN_IRQ_MASK_OFS); ZERO(MV_PCI_SERR_MASK); ZERO(hpriv->irq_cause_ofs); @@ -2045,10 +2390,10 @@ static void mv6_reset_flash(struct mv_host_priv *hpriv, void __iomem *mmio) mv5_reset_flash(hpriv, mmio); - tmp = readl(mmio + MV_GPIO_PORT_CTL); + tmp = readl(mmio + MV_GPIO_PORT_CTL_OFS); tmp &= 0x3; tmp |= (1 << 5) | (1 << 6); - writel(tmp, mmio + MV_GPIO_PORT_CTL); + writel(tmp, mmio + MV_GPIO_PORT_CTL_OFS); } /** @@ -2121,7 +2466,7 @@ static void mv6_read_preamp(struct mv_host_priv *hpriv, int idx, void __iomem *port_mmio; u32 tmp; - tmp = readl(mmio + MV_RESET_CFG); + tmp = readl(mmio + MV_RESET_CFG_OFS); if ((tmp & (1 << 0)) == 0) { hpriv->signal[idx].amps = 0x7 << 8; hpriv->signal[idx].pre = 0x1 << 5; @@ -2137,7 +2482,7 @@ static void mv6_read_preamp(struct mv_host_priv *hpriv, int idx, static void mv6_enable_leds(struct mv_host_priv *hpriv, void __iomem *mmio) { - writel(0x00000060, mmio + MV_GPIO_PORT_CTL); + writel(0x00000060, mmio + MV_GPIO_PORT_CTL_OFS); } static void mv6_phy_errata(struct mv_host_priv *hpriv, void __iomem *mmio, @@ -2235,11 +2580,6 @@ static void mv_soc_reset_hc_port(struct mv_host_priv *hpriv, { void __iomem *port_mmio = mv_port_base(mmio, port); - /* - * The datasheet warns against setting ATA_RST when EDMA is active - * (but doesn't say what the problem might be). So we first try - * to disable the EDMA engine before doing the ATA_RST operation. - */ mv_reset_channel(hpriv, mmio, port); ZERO(0x028); /* command */ @@ -2254,7 +2594,7 @@ static void mv_soc_reset_hc_port(struct mv_host_priv *hpriv, ZERO(0x024); /* respq outp */ ZERO(0x020); /* respq inp */ ZERO(0x02c); /* test control */ - writel(0xbc, port_mmio + EDMA_IORDY_TMOUT); + writel(0xbc, port_mmio + EDMA_IORDY_TMOUT_OFS); } #undef ZERO @@ -2297,38 +2637,39 @@ static void mv_soc_reset_bus(struct ata_host *host, void __iomem *mmio) return; } -static void mv_setup_ifctl(void __iomem *port_mmio, int want_gen2i) +static void mv_setup_ifcfg(void __iomem *port_mmio, int want_gen2i) { - u32 ifctl = readl(port_mmio + SATA_INTERFACE_CFG); + u32 ifcfg = readl(port_mmio + SATA_INTERFACE_CFG_OFS); - ifctl = (ifctl & 0xf7f) | 0x9b1000; /* from chip spec */ + ifcfg = (ifcfg & 0xf7f) | 0x9b1000; /* from chip spec */ if (want_gen2i) - ifctl |= (1 << 7); /* enable gen2i speed */ - writelfl(ifctl, port_mmio + SATA_INTERFACE_CFG); + ifcfg |= (1 << 7); /* enable gen2i speed */ + writelfl(ifcfg, port_mmio + SATA_INTERFACE_CFG_OFS); } -/* - * Caller must ensure that EDMA is not active, - * by first doing mv_stop_edma() where needed. - */ static void mv_reset_channel(struct mv_host_priv *hpriv, void __iomem *mmio, unsigned int port_no) { void __iomem *port_mmio = mv_port_base(mmio, port_no); + /* + * The datasheet warns against setting EDMA_RESET when EDMA is active + * (but doesn't say what the problem might be). So we first try + * to disable the EDMA engine before doing the EDMA_RESET operation. + */ mv_stop_edma_engine(port_mmio); - writelfl(ATA_RST, port_mmio + EDMA_CMD_OFS); + writelfl(EDMA_RESET, port_mmio + EDMA_CMD_OFS); if (!IS_GEN_I(hpriv)) { - /* Enable 3.0gb/s link speed */ - mv_setup_ifctl(port_mmio, 1); + /* Enable 3.0gb/s link speed: this survives EDMA_RESET */ + mv_setup_ifcfg(port_mmio, 1); } /* - * Strobing ATA_RST here causes a hard reset of the SATA transport, + * Strobing EDMA_RESET here causes a hard reset of the SATA transport, * link, and physical layers. It resets all SATA interface registers * (except for SATA_INTERFACE_CFG), and issues a COMRESET to the dev. */ - writelfl(ATA_RST, port_mmio + EDMA_CMD_OFS); + writelfl(EDMA_RESET, port_mmio + EDMA_CMD_OFS); udelay(25); /* allow reset propagation */ writelfl(0, port_mmio + EDMA_CMD_OFS); @@ -2392,7 +2733,7 @@ static int mv_hardreset(struct ata_link *link, unsigned int *class, sata_scr_read(link, SCR_STATUS, &sstatus); if (!IS_GEN_I(hpriv) && ++attempts >= 5 && sstatus == 0x121) { /* Force 1.5gb/s link speed and try again */ - mv_setup_ifctl(mv_ap_base(ap), 0); + mv_setup_ifcfg(mv_ap_base(ap), 0); if (time_after(jiffies + HZ, deadline)) extra = HZ; /* only extend it once, max */ } @@ -2493,6 +2834,34 @@ static void mv_port_init(struct ata_ioports *port, void __iomem *port_mmio) readl(port_mmio + EDMA_ERR_IRQ_MASK_OFS)); } +static unsigned int mv_in_pcix_mode(struct ata_host *host) +{ + struct mv_host_priv *hpriv = host->private_data; + void __iomem *mmio = hpriv->base; + u32 reg; + + if (!HAS_PCI(host) || !IS_PCIE(hpriv)) + return 0; /* not PCI-X capable */ + reg = readl(mmio + MV_PCI_MODE_OFS); + if ((reg & MV_PCI_MODE_MASK) == 0) + return 0; /* conventional PCI mode */ + return 1; /* chip is in PCI-X mode */ +} + +static int mv_pci_cut_through_okay(struct ata_host *host) +{ + struct mv_host_priv *hpriv = host->private_data; + void __iomem *mmio = hpriv->base; + u32 reg; + + if (!mv_in_pcix_mode(host)) { + reg = readl(mmio + PCI_COMMAND_OFS); + if (reg & PCI_COMMAND_MRDTRIG) + return 0; /* not okay */ + } + return 1; /* okay */ +} + static int mv_chip_id(struct ata_host *host, unsigned int board_idx) { struct pci_dev *pdev = to_pci_dev(host->dev); @@ -2560,7 +2929,7 @@ static int mv_chip_id(struct ata_host *host, unsigned int board_idx) break; case chip_7042: - hp_flags |= MV_HP_PCIE; + hp_flags |= MV_HP_PCIE | MV_HP_CUT_THROUGH; if (pdev->vendor == PCI_VENDOR_ID_TTI && (pdev->device == 0x2300 || pdev->device == 0x2310)) { @@ -2590,9 +2959,12 @@ static int mv_chip_id(struct ata_host *host, unsigned int board_idx) " and avoid the final two gigabytes on" " all RocketRAID BIOS initialized drives.\n"); } + /* drop through */ case chip_6042: hpriv->ops = &mv6xxx_ops; hp_flags |= MV_HP_GEN_IIE; + if (board_idx == chip_6042 && mv_pci_cut_through_okay(host)) + hp_flags |= MV_HP_CUT_THROUGH; switch (pdev->revision) { case 0x0: diff --git a/drivers/base/sys.c b/drivers/base/sys.c index 4fbb56bcb1e..358bb0be3c0 100644 --- a/drivers/base/sys.c +++ b/drivers/base/sys.c @@ -175,8 +175,7 @@ int sysdev_driver_register(struct sysdev_class *cls, struct sysdev_driver *drv) } /* Check whether this driver has already been added to a class. */ - if ((drv->entry.next != drv->entry.prev) || - (drv->entry.next != NULL)) { + if (drv->entry.next && !list_empty(&drv->entry)) { printk(KERN_WARNING "sysdev: class %s: driver (%p) has already" " been registered to a class, something is wrong, but " "will forge on!\n", cls->name, drv); diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c index 8fc429cf82b..41f818be2f7 100644 --- a/drivers/block/aoe/aoecmd.c +++ b/drivers/block/aoe/aoecmd.c @@ -755,11 +755,13 @@ diskstats(struct gendisk *disk, struct bio *bio, ulong duration, sector_t sector { unsigned long n_sect = bio->bi_size >> 9; const int rw = bio_data_dir(bio); + struct hd_struct *part; - all_stat_inc(disk, ios[rw], sector); - all_stat_add(disk, ticks[rw], duration, sector); - all_stat_add(disk, sectors[rw], n_sect, sector); - all_stat_add(disk, io_ticks, duration, sector); + part = get_part(disk, sector); + all_stat_inc(disk, part, ios[rw], sector); + all_stat_add(disk, part, ticks[rw], duration, sector); + all_stat_add(disk, part, sectors[rw], n_sect, sector); + all_stat_add(disk, part, io_ticks, duration, sector); } void diff --git a/drivers/char/sx.c b/drivers/char/sx.c index f39f6fd8935..b1a7a8cb65e 100644 --- a/drivers/char/sx.c +++ b/drivers/char/sx.c @@ -970,7 +970,8 @@ static int sx_set_real_termios(void *ptr) sx_write_channel_byte(port, hi_mask, 0x1f); break; default: - printk(KERN_INFO "sx: Invalid wordsize: %u\n", CFLAG & CSIZE); + printk(KERN_INFO "sx: Invalid wordsize: %u\n", + (unsigned int)CFLAG & CSIZE); break; } @@ -997,7 +998,8 @@ static int sx_set_real_termios(void *ptr) set_bit(TTY_HW_COOK_IN, &port->gs.tty->flags); } sx_dprintk(SX_DEBUG_TERMIOS, "iflags: %x(%d) ", - port->gs.tty->termios->c_iflag, I_OTHER(port->gs.tty)); + (unsigned int)port->gs.tty->termios->c_iflag, + I_OTHER(port->gs.tty)); /* Tell line discipline whether we will do output cooking. * If OPOST is set and no other output flags are set then we can do output @@ -1010,7 +1012,8 @@ static int sx_set_real_termios(void *ptr) clear_bit(TTY_HW_COOK_OUT, &port->gs.tty->flags); } sx_dprintk(SX_DEBUG_TERMIOS, "oflags: %x(%d)\n", - port->gs.tty->termios->c_oflag, O_OTHER(port->gs.tty)); + (unsigned int)port->gs.tty->termios->c_oflag, + O_OTHER(port->gs.tty)); /* port->c_dcd = sx_get_CD (port); */ func_exit(); return 0; diff --git a/drivers/char/vt.c b/drivers/char/vt.c index e458b08139a..fa1ffbf2c62 100644 --- a/drivers/char/vt.c +++ b/drivers/char/vt.c @@ -2742,6 +2742,10 @@ static int con_open(struct tty_struct *tty, struct file *filp) tty->winsize.ws_row = vc_cons[currcons].d->vc_rows; tty->winsize.ws_col = vc_cons[currcons].d->vc_cols; } + if (vc->vc_utf) + tty->termios->c_iflag |= IUTF8; + else + tty->termios->c_iflag &= ~IUTF8; release_console_sem(); vcs_make_sysfs(tty); return ret; @@ -2918,6 +2922,8 @@ int __init vty_init(void) console_driver->minor_start = 1; console_driver->type = TTY_DRIVER_TYPE_CONSOLE; console_driver->init_termios = tty_std_termios; + if (default_utf8) + console_driver->init_termios.c_iflag |= IUTF8; console_driver->flags = TTY_DRIVER_REAL_RAW | TTY_DRIVER_RESET_TERMIOS; tty_set_operations(console_driver, &con_ops); if (tty_register_driver(console_driver)) diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.c b/drivers/infiniband/hw/cxgb3/cxio_hal.c index 5fd8506a865..ebf9d3043f8 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.c +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.c @@ -588,7 +588,7 @@ static int cxio_hal_destroy_ctrl_qp(struct cxio_rdev *rdev_p) * caller aquires the ctrl_qp lock before the call */ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, - u32 len, void *data, int completion) + u32 len, void *data) { u32 i, nr_wqe, copy_len; u8 *copy_data; @@ -624,7 +624,7 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, flag = 0; if (i == (nr_wqe - 1)) { /* last WQE */ - flag = completion ? T3_COMPLETION_FLAG : 0; + flag = T3_COMPLETION_FLAG; if (len % 32) utx_len = len / 32 + 1; else @@ -683,21 +683,20 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, return 0; } -/* IN: stag key, pdid, perm, zbva, to, len, page_size, pbl, and pbl_size - * OUT: stag index, actual pbl_size, pbl_addr allocated. +/* IN: stag key, pdid, perm, zbva, to, len, page_size, pbl_size and pbl_addr + * OUT: stag index * TBD: shared memory region support */ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, u32 *stag, u8 stag_state, u32 pdid, enum tpt_mem_type type, enum tpt_mem_perm perm, - u32 zbva, u64 to, u32 len, u8 page_size, __be64 *pbl, - u32 *pbl_size, u32 *pbl_addr) + u32 zbva, u64 to, u32 len, u8 page_size, + u32 pbl_size, u32 pbl_addr) { int err; struct tpt_entry tpt; u32 stag_idx; u32 wptr; - int rereg = (*stag != T3_STAG_UNSET); stag_state = stag_state > 0; stag_idx = (*stag) >> 8; @@ -711,30 +710,8 @@ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, PDBG("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", __func__, stag_state, type, pdid, stag_idx); - if (reset_tpt_entry) - cxio_hal_pblpool_free(rdev_p, *pbl_addr, *pbl_size << 3); - else if (!rereg) { - *pbl_addr = cxio_hal_pblpool_alloc(rdev_p, *pbl_size << 3); - if (!*pbl_addr) { - return -ENOMEM; - } - } - mutex_lock(&rdev_p->ctrl_qp.lock); - /* write PBL first if any - update pbl only if pbl list exist */ - if (pbl) { - - PDBG("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", - __func__, *pbl_addr, rdev_p->rnic_info.pbl_base, - *pbl_size); - err = cxio_hal_ctrl_qp_write_mem(rdev_p, - (*pbl_addr >> 5), - (*pbl_size << 3), pbl, 0); - if (err) - goto ret; - } - /* write TPT entry */ if (reset_tpt_entry) memset(&tpt, 0, sizeof(tpt)); @@ -749,23 +726,23 @@ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, V_TPT_ADDR_TYPE((zbva ? TPT_ZBTO : TPT_VATO)) | V_TPT_PAGE_SIZE(page_size)); tpt.rsvd_pbl_addr = reset_tpt_entry ? 0 : - cpu_to_be32(V_TPT_PBL_ADDR(PBL_OFF(rdev_p, *pbl_addr)>>3)); + cpu_to_be32(V_TPT_PBL_ADDR(PBL_OFF(rdev_p, pbl_addr)>>3)); tpt.len = cpu_to_be32(len); tpt.va_hi = cpu_to_be32((u32) (to >> 32)); tpt.va_low_or_fbo = cpu_to_be32((u32) (to & 0xFFFFFFFFULL)); tpt.rsvd_bind_cnt_or_pstag = 0; tpt.rsvd_pbl_size = reset_tpt_entry ? 0 : - cpu_to_be32(V_TPT_PBL_SIZE((*pbl_size) >> 2)); + cpu_to_be32(V_TPT_PBL_SIZE(pbl_size >> 2)); } err = cxio_hal_ctrl_qp_write_mem(rdev_p, stag_idx + (rdev_p->rnic_info.tpt_base >> 5), - sizeof(tpt), &tpt, 1); + sizeof(tpt), &tpt); /* release the stag index to free pool */ if (reset_tpt_entry) cxio_hal_put_stag(rdev_p->rscp, stag_idx); -ret: + wptr = rdev_p->ctrl_qp.wptr; mutex_unlock(&rdev_p->ctrl_qp.lock); if (!err) @@ -776,44 +753,67 @@ ret: return err; } +int cxio_write_pbl(struct cxio_rdev *rdev_p, __be64 *pbl, + u32 pbl_addr, u32 pbl_size) +{ + u32 wptr; + int err; + + PDBG("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", + __func__, pbl_addr, rdev_p->rnic_info.pbl_base, + pbl_size); + + mutex_lock(&rdev_p->ctrl_qp.lock); + err = cxio_hal_ctrl_qp_write_mem(rdev_p, pbl_addr >> 5, pbl_size << 3, + pbl); + wptr = rdev_p->ctrl_qp.wptr; + mutex_unlock(&rdev_p->ctrl_qp.lock); + if (err) + return err; + + if (wait_event_interruptible(rdev_p->ctrl_qp.waitq, + SEQ32_GE(rdev_p->ctrl_qp.rptr, + wptr))) + return -ERESTARTSYS; + + return 0; +} + int cxio_register_phys_mem(struct cxio_rdev *rdev_p, u32 *stag, u32 pdid, enum tpt_mem_perm perm, u32 zbva, u64 to, u32 len, - u8 page_size, __be64 *pbl, u32 *pbl_size, - u32 *pbl_addr) + u8 page_size, u32 pbl_size, u32 pbl_addr) { *stag = T3_STAG_UNSET; return __cxio_tpt_op(rdev_p, 0, stag, 1, pdid, TPT_NON_SHARED_MR, perm, - zbva, to, len, page_size, pbl, pbl_size, pbl_addr); + zbva, to, len, page_size, pbl_size, pbl_addr); } int cxio_reregister_phys_mem(struct cxio_rdev *rdev_p, u32 *stag, u32 pdid, enum tpt_mem_perm perm, u32 zbva, u64 to, u32 len, - u8 page_size, __be64 *pbl, u32 *pbl_size, - u32 *pbl_addr) + u8 page_size, u32 pbl_size, u32 pbl_addr) { return __cxio_tpt_op(rdev_p, 0, stag, 1, pdid, TPT_NON_SHARED_MR, perm, - zbva, to, len, page_size, pbl, pbl_size, pbl_addr); + zbva, to, len, page_size, pbl_size, pbl_addr); } int cxio_dereg_mem(struct cxio_rdev *rdev_p, u32 stag, u32 pbl_size, u32 pbl_addr) { - return __cxio_tpt_op(rdev_p, 1, &stag, 0, 0, 0, 0, 0, 0ULL, 0, 0, NULL, - &pbl_size, &pbl_addr); + return __cxio_tpt_op(rdev_p, 1, &stag, 0, 0, 0, 0, 0, 0ULL, 0, 0, + pbl_size, pbl_addr); } int cxio_allocate_window(struct cxio_rdev *rdev_p, u32 * stag, u32 pdid) { - u32 pbl_size = 0; *stag = T3_STAG_UNSET; return __cxio_tpt_op(rdev_p, 0, stag, 0, pdid, TPT_MW, 0, 0, 0ULL, 0, 0, - NULL, &pbl_size, NULL); + 0, 0); } int cxio_deallocate_window(struct cxio_rdev *rdev_p, u32 stag) { - return __cxio_tpt_op(rdev_p, 1, &stag, 0, 0, 0, 0, 0, 0ULL, 0, 0, NULL, - NULL, NULL); + return __cxio_tpt_op(rdev_p, 1, &stag, 0, 0, 0, 0, 0, 0ULL, 0, 0, + 0, 0); } int cxio_rdma_init(struct cxio_rdev *rdev_p, struct t3_rdma_init_attr *attr) diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.h b/drivers/infiniband/hw/cxgb3/cxio_hal.h index 69ab08ebc68..6e128f6bab0 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.h +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.h @@ -154,14 +154,14 @@ int cxio_create_qp(struct cxio_rdev *rdev, u32 kernel_domain, struct t3_wq *wq, int cxio_destroy_qp(struct cxio_rdev *rdev, struct t3_wq *wq, struct cxio_ucontext *uctx); int cxio_peek_cq(struct t3_wq *wr, struct t3_cq *cq, int opcode); +int cxio_write_pbl(struct cxio_rdev *rdev_p, __be64 *pbl, + u32 pbl_addr, u32 pbl_size); int cxio_register_phys_mem(struct cxio_rdev *rdev, u32 * stag, u32 pdid, enum tpt_mem_perm perm, u32 zbva, u64 to, u32 len, - u8 page_size, __be64 *pbl, u32 *pbl_size, - u32 *pbl_addr); + u8 page_size, u32 pbl_size, u32 pbl_addr); int cxio_reregister_phys_mem(struct cxio_rdev *rdev, u32 * stag, u32 pdid, enum tpt_mem_perm perm, u32 zbva, u64 to, u32 len, - u8 page_size, __be64 *pbl, u32 *pbl_size, - u32 *pbl_addr); + u8 page_size, u32 pbl_size, u32 pbl_addr); int cxio_dereg_mem(struct cxio_rdev *rdev, u32 stag, u32 pbl_size, u32 pbl_addr); int cxio_allocate_window(struct cxio_rdev *rdev, u32 * stag, u32 pdid); diff --git a/drivers/infiniband/hw/cxgb3/cxio_resource.c b/drivers/infiniband/hw/cxgb3/cxio_resource.c index 45ed4f25ef7..bd233c08765 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_resource.c +++ b/drivers/infiniband/hw/cxgb3/cxio_resource.c @@ -250,7 +250,6 @@ void cxio_hal_destroy_resource(struct cxio_hal_resource *rscp) */ #define MIN_PBL_SHIFT 8 /* 256B == min PBL size (32 entries) */ -#define PBL_CHUNK 2*1024*1024 u32 cxio_hal_pblpool_alloc(struct cxio_rdev *rdev_p, int size) { @@ -267,14 +266,35 @@ void cxio_hal_pblpool_free(struct cxio_rdev *rdev_p, u32 addr, int size) int cxio_hal_pblpool_create(struct cxio_rdev *rdev_p) { - unsigned long i; + unsigned pbl_start, pbl_chunk; + rdev_p->pbl_pool = gen_pool_create(MIN_PBL_SHIFT, -1); - if (rdev_p->pbl_pool) - for (i = rdev_p->rnic_info.pbl_base; - i <= rdev_p->rnic_info.pbl_top - PBL_CHUNK + 1; - i += PBL_CHUNK) - gen_pool_add(rdev_p->pbl_pool, i, PBL_CHUNK, -1); - return rdev_p->pbl_pool ? 0 : -ENOMEM; + if (!rdev_p->pbl_pool) + return -ENOMEM; + + pbl_start = rdev_p->rnic_info.pbl_base; + pbl_chunk = rdev_p->rnic_info.pbl_top - pbl_start + 1; + + while (pbl_start < rdev_p->rnic_info.pbl_top) { + pbl_chunk = min(rdev_p->rnic_info.pbl_top - pbl_start + 1, + pbl_chunk); + if (gen_pool_add(rdev_p->pbl_pool, pbl_start, pbl_chunk, -1)) { + PDBG("%s failed to add PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); + if (pbl_chunk <= 1024 << MIN_PBL_SHIFT) { + printk(KERN_WARNING MOD "%s: Failed to add all PBL chunks (%x/%x)\n", + __func__, pbl_start, rdev_p->rnic_info.pbl_top - pbl_start); + return 0; + } + pbl_chunk >>= 1; + } else { + PDBG("%s added PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); + pbl_start += pbl_chunk; + } + } + + return 0; } void cxio_hal_pblpool_destroy(struct cxio_rdev *rdev_p) diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c index 58c3d61bcd1..ec49a5cbdeb 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_mem.c +++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c @@ -35,17 +35,26 @@ #include <rdma/ib_verbs.h> #include "cxio_hal.h" +#include "cxio_resource.h" #include "iwch.h" #include "iwch_provider.h" -int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php, - struct iwch_mr *mhp, - int shift, - __be64 *page_list) +static void iwch_finish_mem_reg(struct iwch_mr *mhp, u32 stag) { - u32 stag; u32 mmid; + mhp->attr.state = 1; + mhp->attr.stag = stag; + mmid = stag >> 8; + mhp->ibmr.rkey = mhp->ibmr.lkey = stag; + insert_handle(mhp->rhp, &mhp->rhp->mmidr, mhp, mmid); + PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); +} + +int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php, + struct iwch_mr *mhp, int shift) +{ + u32 stag; if (cxio_register_phys_mem(&rhp->rdev, &stag, mhp->attr.pdid, @@ -53,28 +62,21 @@ int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php, mhp->attr.zbva, mhp->attr.va_fbo, mhp->attr.len, - shift-12, - page_list, - &mhp->attr.pbl_size, &mhp->attr.pbl_addr)) + shift - 12, + mhp->attr.pbl_size, mhp->attr.pbl_addr)) return -ENOMEM; - mhp->attr.state = 1; - mhp->attr.stag = stag; - mmid = stag >> 8; - mhp->ibmr.rkey = mhp->ibmr.lkey = stag; - insert_handle(rhp, &rhp->mmidr, mhp, mmid); - PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); + + iwch_finish_mem_reg(mhp, stag); + return 0; } int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php, struct iwch_mr *mhp, int shift, - __be64 *page_list, int npages) { u32 stag; - u32 mmid; - /* We could support this... */ if (npages > mhp->attr.pbl_size) @@ -87,19 +89,40 @@ int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php, mhp->attr.zbva, mhp->attr.va_fbo, mhp->attr.len, - shift-12, - page_list, - &mhp->attr.pbl_size, &mhp->attr.pbl_addr)) + shift - 12, + mhp->attr.pbl_size, mhp->attr.pbl_addr)) return -ENOMEM; - mhp->attr.state = 1; - mhp->attr.stag = stag; - mmid = stag >> 8; - mhp->ibmr.rkey = mhp->ibmr.lkey = stag; - insert_handle(rhp, &rhp->mmidr, mhp, mmid); - PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); + + iwch_finish_mem_reg(mhp, stag); + + return 0; +} + +int iwch_alloc_pbl(struct iwch_mr *mhp, int npages) +{ + mhp->attr.pbl_addr = cxio_hal_pblpool_alloc(&mhp->rhp->rdev, + npages << 3); + + if (!mhp->attr.pbl_addr) + return -ENOMEM; + + mhp->attr.pbl_size = npages; + return 0; } +void iwch_free_pbl(struct iwch_mr *mhp) +{ + cxio_hal_pblpool_free(&mhp->rhp->rdev, mhp->attr.pbl_addr, + mhp->attr.pbl_size << 3); +} + +int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset) +{ + return cxio_write_pbl(&mhp->rhp->rdev, pages, + mhp->attr.pbl_addr + (offset << 3), npages); +} + int build_phys_page_list(struct ib_phys_buf *buffer_list, int num_phys_buf, u64 *iova_start, diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c index d07d3a377b5..8934178a23e 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.c +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c @@ -442,6 +442,7 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr) mmid = mhp->attr.stag >> 8; cxio_dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size, mhp->attr.pbl_addr); + iwch_free_pbl(mhp); remove_handle(rhp, &rhp->mmidr, mmid); if (mhp->kva) kfree((void *) (unsigned long) mhp->kva); @@ -475,6 +476,8 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd, if (!mhp) return ERR_PTR(-ENOMEM); + mhp->rhp = rhp; + /* First check that we have enough alignment */ if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK)) { ret = -EINVAL; @@ -492,7 +495,17 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd, if (ret) goto err; - mhp->rhp = rhp; + ret = iwch_alloc_pbl(mhp, npages); + if (ret) { + kfree(page_list); + goto err_pbl; + } + + ret = iwch_write_pbl(mhp, page_list, npages, 0); + kfree(page_list); + if (ret) + goto err_pbl; + mhp->attr.pdid = php->pdid; mhp->attr.zbva = 0; @@ -502,12 +515,15 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd, mhp->attr.len = (u32) total_size; mhp->attr.pbl_size = npages; - ret = iwch_register_mem(rhp, php, mhp, shift, page_list); - kfree(page_list); - if (ret) { - goto err; - } + ret = iwch_register_mem(rhp, php, mhp, shift); + if (ret) + goto err_pbl; + return &mhp->ibmr; + +err_pbl: + iwch_free_pbl(mhp); + err: kfree(mhp); return ERR_PTR(ret); @@ -560,7 +576,7 @@ static int iwch_reregister_phys_mem(struct ib_mr *mr, return ret; } - ret = iwch_reregister_mem(rhp, php, &mh, shift, page_list, npages); + ret = iwch_reregister_mem(rhp, php, &mh, shift, npages); kfree(page_list); if (ret) { return ret; @@ -602,6 +618,8 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, if (!mhp) return ERR_PTR(-ENOMEM); + mhp->rhp = rhp; + mhp->umem = ib_umem_get(pd->uobject->context, start, length, acc, 0); if (IS_ERR(mhp->umem)) { err = PTR_ERR(mhp->umem); @@ -615,10 +633,14 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, list_for_each_entry(chunk, &mhp->umem->chunk_list, list) n += chunk->nents; - pages = kmalloc(n * sizeof(u64), GFP_KERNEL); + err = iwch_alloc_pbl(mhp, n); + if (err) + goto err; + + pages = (__be64 *) __get_free_page(GFP_KERNEL); if (!pages) { err = -ENOMEM; - goto err; + goto err_pbl; } i = n = 0; @@ -630,25 +652,38 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, pages[i++] = cpu_to_be64(sg_dma_address( &chunk->page_list[j]) + mhp->umem->page_size * k); + if (i == PAGE_SIZE / sizeof *pages) { + err = iwch_write_pbl(mhp, pages, i, n); + if (err) + goto pbl_done; + n += i; + i = 0; + } } } - mhp->rhp = rhp; + if (i) + err = iwch_write_pbl(mhp, pages, i, n); + +pbl_done: + free_page((unsigned long) pages); + if (err) + goto err_pbl; + mhp->attr.pdid = php->pdid; mhp->attr.zbva = 0; mhp->attr.perms = iwch_ib_to_tpt_access(acc); mhp->attr.va_fbo = virt; mhp->attr.page_size = shift - 12; mhp->attr.len = (u32) length; - mhp->attr.pbl_size = i; - err = iwch_register_mem(rhp, php, mhp, shift, pages); - kfree(pages); + + err = iwch_register_mem(rhp, php, mhp, shift); if (err) - goto err; + goto err_pbl; if (udata && !t3a_device(rhp)) { uresp.pbl_addr = (mhp->attr.pbl_addr - - rhp->rdev.rnic_info.pbl_base) >> 3; + rhp->rdev.rnic_info.pbl_base) >> 3; PDBG("%s user resp pbl_addr 0x%x\n", __func__, uresp.pbl_addr); @@ -661,6 +696,9 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, return &mhp->ibmr; +err_pbl: + iwch_free_pbl(mhp); + err: ib_umem_release(mhp->umem); kfree(mhp); diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h index db5100d27ca..836163fc542 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.h +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h @@ -340,14 +340,14 @@ int iwch_quiesce_qps(struct iwch_cq *chp); int iwch_resume_qps(struct iwch_cq *chp); void stop_read_rep_timer(struct iwch_qp *qhp); int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php, - struct iwch_mr *mhp, - int shift, - __be64 *page_list); + struct iwch_mr *mhp, int shift); int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php, struct iwch_mr *mhp, int shift, - __be64 *page_list, int npages); +int iwch_alloc_pbl(struct iwch_mr *mhp, int npages); +void iwch_free_pbl(struct iwch_mr *mhp); +int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset); int build_phys_page_list(struct ib_phys_buf *buffer_list, int num_phys_buf, u64 *iova_start, diff --git a/drivers/infiniband/hw/ehca/ehca_classes.h b/drivers/infiniband/hw/ehca/ehca_classes.h index 00bab60f6de..1e9e99a1393 100644 --- a/drivers/infiniband/hw/ehca/ehca_classes.h +++ b/drivers/infiniband/hw/ehca/ehca_classes.h @@ -192,6 +192,8 @@ struct ehca_qp { int mtu_shift; u32 message_count; u32 packet_count; + atomic_t nr_events; /* events seen */ + wait_queue_head_t wait_completion; }; #define IS_SRQ(qp) (qp->ext_type == EQPT_SRQ) diff --git a/drivers/infiniband/hw/ehca/ehca_irq.c b/drivers/infiniband/hw/ehca/ehca_irq.c index ca5eb0cb628..ce1ab0571be 100644 --- a/drivers/infiniband/hw/ehca/ehca_irq.c +++ b/drivers/infiniband/hw/ehca/ehca_irq.c @@ -204,6 +204,8 @@ static void qp_event_callback(struct ehca_shca *shca, u64 eqe, read_lock(&ehca_qp_idr_lock); qp = idr_find(&ehca_qp_idr, token); + if (qp) + atomic_inc(&qp->nr_events); read_unlock(&ehca_qp_idr_lock); if (!qp) @@ -223,6 +225,8 @@ static void qp_event_callback(struct ehca_shca *shca, u64 eqe, if (fatal && qp->ext_type == EQPT_SRQBASE) dispatch_qp_event(shca, qp, IB_EVENT_QP_LAST_WQE_REACHED); + if (atomic_dec_and_test(&qp->nr_events)) + wake_up(&qp->wait_completion); return; } diff --git a/drivers/infiniband/hw/ehca/ehca_qp.c b/drivers/infiniband/hw/ehca/ehca_qp.c index 18fba92fa7a..3f59587338e 100644 --- a/drivers/infiniband/hw/ehca/ehca_qp.c +++ b/drivers/infiniband/hw/ehca/ehca_qp.c @@ -566,6 +566,8 @@ static struct ehca_qp *internal_create_qp( return ERR_PTR(-ENOMEM); } + atomic_set(&my_qp->nr_events, 0); + init_waitqueue_head(&my_qp->wait_completion); spin_lock_init(&my_qp->spinlock_s); spin_lock_init(&my_qp->spinlock_r); my_qp->qp_type = qp_type; @@ -1934,6 +1936,9 @@ static int internal_destroy_qp(struct ib_device *dev, struct ehca_qp *my_qp, idr_remove(&ehca_qp_idr, my_qp->token); write_unlock_irqrestore(&ehca_qp_idr_lock, flags); + /* now wait until all pending events have completed */ + wait_event(my_qp->wait_completion, !atomic_read(&my_qp->nr_events)); + h_ret = hipz_h_destroy_qp(shca->ipz_hca_handle, my_qp); if (h_ret != H_SUCCESS) { ehca_err(dev, "hipz_h_destroy_qp() failed h_ret=%li " diff --git a/drivers/infiniband/hw/ipath/ipath_driver.c b/drivers/infiniband/hw/ipath/ipath_driver.c index acf30c06a0c..ce7b7c34360 100644 --- a/drivers/infiniband/hw/ipath/ipath_driver.c +++ b/drivers/infiniband/hw/ipath/ipath_driver.c @@ -1197,7 +1197,7 @@ void ipath_kreceive(struct ipath_portdata *pd) } reloop: - for (last = 0, i = 1; !last; i++) { + for (last = 0, i = 1; !last; i += !last) { hdr = dd->ipath_f_get_msgheader(dd, rhf_addr); eflags = ipath_hdrget_err_flags(rhf_addr); etype = ipath_hdrget_rcv_type(rhf_addr); @@ -1428,6 +1428,40 @@ static void ipath_update_pio_bufs(struct ipath_devdata *dd) spin_unlock_irqrestore(&ipath_pioavail_lock, flags); } +/* + * used to force update of pioavailshadow if we can't get a pio buffer. + * Needed primarily due to exitting freeze mode after recovering + * from errors. Done lazily, because it's safer (known to not + * be writing pio buffers). + */ +static void ipath_reset_availshadow(struct ipath_devdata *dd) +{ + int i, im; + unsigned long flags; + + spin_lock_irqsave(&ipath_pioavail_lock, flags); + for (i = 0; i < dd->ipath_pioavregs; i++) { + u64 val, oldval; + /* deal with 6110 chip bug on high register #s */ + im = (i > 3 && (dd->ipath_flags & IPATH_SWAP_PIOBUFS)) ? + i ^ 1 : i; + val = le64_to_cpu(dd->ipath_pioavailregs_dma[im]); + /* + * busy out the buffers not in the kernel avail list, + * without changing the generation bits. + */ + oldval = dd->ipath_pioavailshadow[i]; + dd->ipath_pioavailshadow[i] = val | + ((~dd->ipath_pioavailkernel[i] << + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT) & + 0xaaaaaaaaaaaaaaaaULL); /* All BUSY bits in qword */ + if (oldval != dd->ipath_pioavailshadow[i]) + ipath_dbg("shadow[%d] was %Lx, now %lx\n", + i, oldval, dd->ipath_pioavailshadow[i]); + } + spin_unlock_irqrestore(&ipath_pioavail_lock, flags); +} + /** * ipath_setrcvhdrsize - set the receive header size * @dd: the infinipath device @@ -1482,9 +1516,12 @@ static noinline void no_pio_bufs(struct ipath_devdata *dd) */ ipath_stats.sps_nopiobufs++; if (!(++dd->ipath_consec_nopiobuf % 100000)) { - ipath_dbg("%u pio sends with no bufavail; dmacopy: " - "%llx %llx %llx %llx; shadow: %lx %lx %lx %lx\n", + ipath_force_pio_avail_update(dd); /* at start */ + ipath_dbg("%u tries no piobufavail ts%lx; dmacopy: " + "%llx %llx %llx %llx\n" + "ipath shadow: %lx %lx %lx %lx\n", dd->ipath_consec_nopiobuf, + (unsigned long)get_cycles(), (unsigned long long) le64_to_cpu(dma[0]), (unsigned long long) le64_to_cpu(dma[1]), (unsigned long long) le64_to_cpu(dma[2]), @@ -1496,14 +1533,17 @@ static noinline void no_pio_bufs(struct ipath_devdata *dd) */ if ((dd->ipath_piobcnt2k + dd->ipath_piobcnt4k) > (sizeof(shadow[0]) * 4 * 4)) - ipath_dbg("2nd group: dmacopy: %llx %llx " - "%llx %llx; shadow: %lx %lx %lx %lx\n", + ipath_dbg("2nd group: dmacopy: " + "%llx %llx %llx %llx\n" + "ipath shadow: %lx %lx %lx %lx\n", (unsigned long long)le64_to_cpu(dma[4]), (unsigned long long)le64_to_cpu(dma[5]), (unsigned long long)le64_to_cpu(dma[6]), (unsigned long long)le64_to_cpu(dma[7]), - shadow[4], shadow[5], shadow[6], - shadow[7]); + shadow[4], shadow[5], shadow[6], shadow[7]); + + /* at end, so update likely happened */ + ipath_reset_availshadow(dd); } } @@ -1652,19 +1692,46 @@ void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start, unsigned len, int avail) { unsigned long flags; - unsigned end; + unsigned end, cnt = 0, next; /* There are two bits per send buffer (busy and generation) */ start *= 2; - len *= 2; - end = start + len; + end = start + len * 2; - /* Set or clear the generation bits. */ spin_lock_irqsave(&ipath_pioavail_lock, flags); + /* Set or clear the busy bit in the shadow. */ while (start < end) { if (avail) { - __clear_bit(start + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT, - dd->ipath_pioavailshadow); + unsigned long dma; + int i, im; + /* + * the BUSY bit will never be set, because we disarm + * the user buffers before we hand them back to the + * kernel. We do have to make sure the generation + * bit is set correctly in shadow, since it could + * have changed many times while allocated to user. + * We can't use the bitmap functions on the full + * dma array because it is always little-endian, so + * we have to flip to host-order first. + * BITS_PER_LONG is slightly wrong, since it's + * always 64 bits per register in chip... + * We only work on 64 bit kernels, so that's OK. + */ + /* deal with 6110 chip bug on high register #s */ + i = start / BITS_PER_LONG; + im = (i > 3 && (dd->ipath_flags & IPATH_SWAP_PIOBUFS)) ? + i ^ 1 : i; + __clear_bit(INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT + + start, dd->ipath_pioavailshadow); + dma = (unsigned long) le64_to_cpu( + dd->ipath_pioavailregs_dma[im]); + if (test_bit((INFINIPATH_SENDPIOAVAIL_CHECK_SHIFT + + start) % BITS_PER_LONG, &dma)) + __set_bit(INFINIPATH_SENDPIOAVAIL_CHECK_SHIFT + + start, dd->ipath_pioavailshadow); + else + __clear_bit(INFINIPATH_SENDPIOAVAIL_CHECK_SHIFT + + start, dd->ipath_pioavailshadow); __set_bit(start, dd->ipath_pioavailkernel); } else { __set_bit(start + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT, @@ -1673,7 +1740,44 @@ void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start, } start += 2; } + + if (dd->ipath_pioupd_thresh) { + end = 2 * (dd->ipath_piobcnt2k + dd->ipath_piobcnt4k); + next = find_first_bit(dd->ipath_pioavailkernel, end); + while (next < end) { + cnt++; + next = find_next_bit(dd->ipath_pioavailkernel, end, + next + 1); + } + } spin_unlock_irqrestore(&ipath_pioavail_lock, flags); + + /* + * When moving buffers from kernel to user, if number assigned to + * the user is less than the pio update threshold, and threshold + * is supported (cnt was computed > 0), drop the update threshold + * so we update at least once per allocated number of buffers. + * In any case, if the kernel buffers are less than the threshold, + * drop the threshold. We don't bother increasing it, having once + * decreased it, since it would typically just cycle back and forth. + * If we don't decrease below buffers in use, we can wait a long + * time for an update, until some other context uses PIO buffers. + */ + if (!avail && len < cnt) + cnt = len; + if (cnt < dd->ipath_pioupd_thresh) { + dd->ipath_pioupd_thresh = cnt; + ipath_dbg("Decreased pio update threshold to %u\n", + dd->ipath_pioupd_thresh); + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl &= ~(INFINIPATH_S_UPDTHRESH_MASK + << INFINIPATH_S_UPDTHRESH_SHIFT); + dd->ipath_sendctrl |= dd->ipath_pioupd_thresh + << INFINIPATH_S_UPDTHRESH_SHIFT; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + } } /** @@ -1794,8 +1898,8 @@ void ipath_cancel_sends(struct ipath_devdata *dd, int restore_sendctrl) spin_lock_irqsave(&dd->ipath_sdma_lock, flags); skip_cancel = - !test_bit(IPATH_SDMA_DISABLED, statp) && - test_and_set_bit(IPATH_SDMA_ABORTING, statp); + test_and_set_bit(IPATH_SDMA_ABORTING, statp) + && !test_bit(IPATH_SDMA_DISABLED, statp); spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); if (skip_cancel) goto bail; @@ -1826,6 +1930,9 @@ void ipath_cancel_sends(struct ipath_devdata *dd, int restore_sendctrl) ipath_disarm_piobufs(dd, 0, dd->ipath_piobcnt2k + dd->ipath_piobcnt4k); + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + set_bit(IPATH_SDMA_DISARMED, &dd->ipath_sdma_status); + if (restore_sendctrl) { /* else done by caller later if needed */ spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); @@ -1845,7 +1952,6 @@ void ipath_cancel_sends(struct ipath_devdata *dd, int restore_sendctrl) /* only wait so long for intr */ dd->ipath_sdma_abort_intr_timeout = jiffies + HZ; dd->ipath_sdma_reset_wait = 200; - __set_bit(IPATH_SDMA_DISARMED, &dd->ipath_sdma_status); if (!test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) tasklet_hi_schedule(&dd->ipath_sdma_abort_task); spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); diff --git a/drivers/infiniband/hw/ipath/ipath_file_ops.c b/drivers/infiniband/hw/ipath/ipath_file_ops.c index 8b1752202e7..3295177c937 100644 --- a/drivers/infiniband/hw/ipath/ipath_file_ops.c +++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c @@ -173,47 +173,25 @@ static int ipath_get_base_info(struct file *fp, (void *) dd->ipath_statusp - (void *) dd->ipath_pioavailregs_dma; if (!shared) { - kinfo->spi_piocnt = dd->ipath_pbufsport; + kinfo->spi_piocnt = pd->port_piocnt; kinfo->spi_piobufbase = (u64) pd->port_piobufs; kinfo->__spi_uregbase = (u64) dd->ipath_uregbase + dd->ipath_ureg_align * pd->port_port; } else if (master) { - kinfo->spi_piocnt = (dd->ipath_pbufsport / subport_cnt) + - (dd->ipath_pbufsport % subport_cnt); + kinfo->spi_piocnt = (pd->port_piocnt / subport_cnt) + + (pd->port_piocnt % subport_cnt); /* Master's PIO buffers are after all the slave's */ kinfo->spi_piobufbase = (u64) pd->port_piobufs + dd->ipath_palign * - (dd->ipath_pbufsport - kinfo->spi_piocnt); + (pd->port_piocnt - kinfo->spi_piocnt); } else { unsigned slave = subport_fp(fp) - 1; - kinfo->spi_piocnt = dd->ipath_pbufsport / subport_cnt; + kinfo->spi_piocnt = pd->port_piocnt / subport_cnt; kinfo->spi_piobufbase = (u64) pd->port_piobufs + dd->ipath_palign * kinfo->spi_piocnt * slave; } - /* - * Set the PIO avail update threshold to no larger - * than the number of buffers per process. Note that - * we decrease it here, but won't ever increase it. - */ - if (dd->ipath_pioupd_thresh && - kinfo->spi_piocnt < dd->ipath_pioupd_thresh) { - unsigned long flags; - - dd->ipath_pioupd_thresh = kinfo->spi_piocnt; - ipath_dbg("Decreased pio update threshold to %u\n", - dd->ipath_pioupd_thresh); - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - dd->ipath_sendctrl &= ~(INFINIPATH_S_UPDTHRESH_MASK - << INFINIPATH_S_UPDTHRESH_SHIFT); - dd->ipath_sendctrl |= dd->ipath_pioupd_thresh - << INFINIPATH_S_UPDTHRESH_SHIFT; - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); - } - if (shared) { kinfo->spi_port_uregbase = (u64) dd->ipath_uregbase + dd->ipath_ureg_align * pd->port_port; @@ -1309,19 +1287,19 @@ static int ipath_mmap(struct file *fp, struct vm_area_struct *vma) ureg = dd->ipath_uregbase + dd->ipath_ureg_align * pd->port_port; if (!pd->port_subport_cnt) { /* port is not shared */ - piocnt = dd->ipath_pbufsport; + piocnt = pd->port_piocnt; piobufs = pd->port_piobufs; } else if (!subport_fp(fp)) { /* caller is the master */ - piocnt = (dd->ipath_pbufsport / pd->port_subport_cnt) + - (dd->ipath_pbufsport % pd->port_subport_cnt); + piocnt = (pd->port_piocnt / pd->port_subport_cnt) + + (pd->port_piocnt % pd->port_subport_cnt); piobufs = pd->port_piobufs + - dd->ipath_palign * (dd->ipath_pbufsport - piocnt); + dd->ipath_palign * (pd->port_piocnt - piocnt); } else { unsigned slave = subport_fp(fp) - 1; /* caller is a slave */ - piocnt = dd->ipath_pbufsport / pd->port_subport_cnt; + piocnt = pd->port_piocnt / pd->port_subport_cnt; piobufs = pd->port_piobufs + dd->ipath_palign * piocnt * slave; } @@ -1633,9 +1611,6 @@ static int try_alloc_port(struct ipath_devdata *dd, int port, port_fp(fp) = pd; pd->port_pid = current->pid; strncpy(pd->port_comm, current->comm, sizeof(pd->port_comm)); - ipath_chg_pioavailkernel(dd, - dd->ipath_pbufsport * (pd->port_port - 1), - dd->ipath_pbufsport, 0); ipath_stats.sps_ports++; ret = 0; } else @@ -1938,11 +1913,25 @@ static int ipath_do_user_init(struct file *fp, /* for now we do nothing with rcvhdrcnt: uinfo->spu_rcvhdrcnt */ + /* some ports may get extra buffers, calculate that here */ + if (pd->port_port <= dd->ipath_ports_extrabuf) + pd->port_piocnt = dd->ipath_pbufsport + 1; + else + pd->port_piocnt = dd->ipath_pbufsport; + /* for right now, kernel piobufs are at end, so port 1 is at 0 */ + if (pd->port_port <= dd->ipath_ports_extrabuf) + pd->port_pio_base = (dd->ipath_pbufsport + 1) + * (pd->port_port - 1); + else + pd->port_pio_base = dd->ipath_ports_extrabuf + + dd->ipath_pbufsport * (pd->port_port - 1); pd->port_piobufs = dd->ipath_piobufbase + - dd->ipath_pbufsport * (pd->port_port - 1) * dd->ipath_palign; - ipath_cdbg(VERBOSE, "Set base of piobufs for port %u to 0x%x\n", - pd->port_port, pd->port_piobufs); + pd->port_pio_base * dd->ipath_palign; + ipath_cdbg(VERBOSE, "piobuf base for port %u is 0x%x, piocnt %u," + " first pio %u\n", pd->port_port, pd->port_piobufs, + pd->port_piocnt, pd->port_pio_base); + ipath_chg_pioavailkernel(dd, pd->port_pio_base, pd->port_piocnt, 0); /* * Now allocate the rcvhdr Q and eager TIDs; skip the TID @@ -2107,7 +2096,6 @@ static int ipath_close(struct inode *in, struct file *fp) } if (dd->ipath_kregbase) { - int i; /* atomically clear receive enable port and intr avail. */ clear_bit(dd->ipath_r_portenable_shift + port, &dd->ipath_rcvctrl); @@ -2136,9 +2124,9 @@ static int ipath_close(struct inode *in, struct file *fp) ipath_write_kreg_port(dd, dd->ipath_kregs->kr_rcvhdraddr, pd->port_port, dd->ipath_dummy_hdrq_phys); - i = dd->ipath_pbufsport * (port - 1); - ipath_disarm_piobufs(dd, i, dd->ipath_pbufsport); - ipath_chg_pioavailkernel(dd, i, dd->ipath_pbufsport, 1); + ipath_disarm_piobufs(dd, pd->port_pio_base, pd->port_piocnt); + ipath_chg_pioavailkernel(dd, pd->port_pio_base, + pd->port_piocnt, 1); dd->ipath_f_clear_tids(dd, pd->port_port); diff --git a/drivers/infiniband/hw/ipath/ipath_iba7220.c b/drivers/infiniband/hw/ipath/ipath_iba7220.c index e3ec0d1bdf5..8eee7830f04 100644 --- a/drivers/infiniband/hw/ipath/ipath_iba7220.c +++ b/drivers/infiniband/hw/ipath/ipath_iba7220.c @@ -595,7 +595,7 @@ static void ipath_7220_txe_recover(struct ipath_devdata *dd) dev_info(&dd->pcidev->dev, "Recovering from TXE PIO parity error\n"); - ipath_disarm_senderrbufs(dd, 1); + ipath_disarm_senderrbufs(dd); } @@ -675,10 +675,8 @@ static void ipath_7220_handle_hwerrors(struct ipath_devdata *dd, char *msg, ctrl = ipath_read_kreg32(dd, dd->ipath_kregs->kr_control); if ((ctrl & INFINIPATH_C_FREEZEMODE) && !ipath_diag_inuse) { /* - * Parity errors in send memory are recoverable, - * just cancel the send (if indicated in * sendbuffererror), - * count the occurrence, unfreeze (if no other handled - * hardware error bits are set), and continue. + * Parity errors in send memory are recoverable by h/w + * just do housekeeping, exit freeze mode and continue. */ if (hwerrs & ((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) @@ -687,13 +685,6 @@ static void ipath_7220_handle_hwerrors(struct ipath_devdata *dd, char *msg, hwerrs &= ~((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT); - if (!hwerrs) { - /* else leave in freeze mode */ - ipath_write_kreg(dd, - dd->ipath_kregs->kr_control, - dd->ipath_control); - goto bail; - } } if (hwerrs) { /* @@ -723,8 +714,8 @@ static void ipath_7220_handle_hwerrors(struct ipath_devdata *dd, char *msg, *dd->ipath_statusp |= IPATH_STATUS_HWERROR; dd->ipath_flags &= ~IPATH_INITTED; } else { - ipath_dbg("Clearing freezemode on ignored hardware " - "error\n"); + ipath_dbg("Clearing freezemode on ignored or " + "recovered hardware error\n"); ipath_clear_freeze(dd); } } @@ -870,8 +861,9 @@ static int ipath_7220_boardname(struct ipath_devdata *dd, char *name, "revision %u.%u!\n", dd->ipath_majrev, dd->ipath_minrev); ret = 1; - } else if (dd->ipath_minrev == 1) { - /* Rev1 chips are prototype. Complain, but allow use */ + } else if (dd->ipath_minrev == 1 && + !(dd->ipath_flags & IPATH_INITTED)) { + /* Rev1 chips are prototype. Complain at init, but allow use */ ipath_dev_err(dd, "Unsupported hardware " "revision %u.%u, Contact support@qlogic.com\n", dd->ipath_majrev, dd->ipath_minrev); @@ -1966,7 +1958,7 @@ static void ipath_7220_config_ports(struct ipath_devdata *dd, ushort cfgports) dd->ipath_rcvctrl); dd->ipath_p0_rcvegrcnt = 2048; /* always */ if (dd->ipath_flags & IPATH_HAS_SEND_DMA) - dd->ipath_pioreserved = 1; /* reserve a buffer */ + dd->ipath_pioreserved = 3; /* kpiobufs used for PIO */ } diff --git a/drivers/infiniband/hw/ipath/ipath_init_chip.c b/drivers/infiniband/hw/ipath/ipath_init_chip.c index 27dd8947666..3e5baa43fc8 100644 --- a/drivers/infiniband/hw/ipath/ipath_init_chip.c +++ b/drivers/infiniband/hw/ipath/ipath_init_chip.c @@ -41,7 +41,7 @@ /* * min buffers we want to have per port, after driver */ -#define IPATH_MIN_USER_PORT_BUFCNT 8 +#define IPATH_MIN_USER_PORT_BUFCNT 7 /* * Number of ports we are configured to use (to allow for more pio @@ -54,13 +54,9 @@ MODULE_PARM_DESC(cfgports, "Set max number of ports to use"); /* * Number of buffers reserved for driver (verbs and layered drivers.) - * Reserved at end of buffer list. Initialized based on - * number of PIO buffers if not set via module interface. + * Initialized based on number of PIO buffers if not set via module interface. * The problem with this is that it's global, but we'll use different - * numbers for different chip types. So the default value is not - * very useful. I've redefined it for the 1.3 release so that it's - * zero unless set by the user to something else, in which case we - * try to respect it. + * numbers for different chip types. */ static ushort ipath_kpiobufs; @@ -546,9 +542,12 @@ static void enable_chip(struct ipath_devdata *dd, int reinit) pioavail = dd->ipath_pioavailregs_dma[i ^ 1]; else pioavail = dd->ipath_pioavailregs_dma[i]; - dd->ipath_pioavailshadow[i] = le64_to_cpu(pioavail) | - (~dd->ipath_pioavailkernel[i] << - INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT); + /* + * don't need to worry about ipath_pioavailkernel here + * because we will call ipath_chg_pioavailkernel() later + * in initialization, to busy out buffers as needed + */ + dd->ipath_pioavailshadow[i] = le64_to_cpu(pioavail); } /* can get counters, stats, etc. */ dd->ipath_flags |= IPATH_PRESENT; @@ -708,12 +707,11 @@ static void verify_interrupt(unsigned long opaque) int ipath_init_chip(struct ipath_devdata *dd, int reinit) { int ret = 0; - u32 val32, kpiobufs; + u32 kpiobufs, defkbufs; u32 piobufs, uports; u64 val; struct ipath_portdata *pd; gfp_t gfp_flags = GFP_USER | __GFP_COMP; - unsigned long flags; ret = init_housekeeping(dd, reinit); if (ret) @@ -753,56 +751,46 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) dd->ipath_pioavregs = ALIGN(piobufs, sizeof(u64) * BITS_PER_BYTE / 2) / (sizeof(u64) * BITS_PER_BYTE / 2); uports = dd->ipath_cfgports ? dd->ipath_cfgports - 1 : 0; - if (ipath_kpiobufs == 0) { - /* not set by user (this is default) */ - if (piobufs > 144) - kpiobufs = 32; - else - kpiobufs = 16; - } + if (piobufs > 144) + defkbufs = 32 + dd->ipath_pioreserved; else - kpiobufs = ipath_kpiobufs; + defkbufs = 16 + dd->ipath_pioreserved; - if (kpiobufs + (uports * IPATH_MIN_USER_PORT_BUFCNT) > piobufs) { + if (ipath_kpiobufs && (ipath_kpiobufs + + (uports * IPATH_MIN_USER_PORT_BUFCNT)) > piobufs) { int i = (int) piobufs - (int) (uports * IPATH_MIN_USER_PORT_BUFCNT); if (i < 1) i = 1; dev_info(&dd->pcidev->dev, "Allocating %d PIO bufs of " "%d for kernel leaves too few for %d user ports " - "(%d each); using %u\n", kpiobufs, + "(%d each); using %u\n", ipath_kpiobufs, piobufs, uports, IPATH_MIN_USER_PORT_BUFCNT, i); /* * shouldn't change ipath_kpiobufs, because could be * different for different devices... */ kpiobufs = i; - } + } else if (ipath_kpiobufs) + kpiobufs = ipath_kpiobufs; + else + kpiobufs = defkbufs; dd->ipath_lastport_piobuf = piobufs - kpiobufs; dd->ipath_pbufsport = uports ? dd->ipath_lastport_piobuf / uports : 0; - val32 = dd->ipath_lastport_piobuf - (dd->ipath_pbufsport * uports); - if (val32 > 0) { - ipath_dbg("allocating %u pbufs/port leaves %u unused, " - "add to kernel\n", dd->ipath_pbufsport, val32); - dd->ipath_lastport_piobuf -= val32; - kpiobufs += val32; - ipath_dbg("%u pbufs/port leaves %u unused, add to kernel\n", - dd->ipath_pbufsport, val32); - } + /* if not an even divisor, some user ports get extra buffers */ + dd->ipath_ports_extrabuf = dd->ipath_lastport_piobuf - + (dd->ipath_pbufsport * uports); + if (dd->ipath_ports_extrabuf) + ipath_dbg("%u pbufs/port leaves some unused, add 1 buffer to " + "ports <= %u\n", dd->ipath_pbufsport, + dd->ipath_ports_extrabuf); dd->ipath_lastpioindex = 0; dd->ipath_lastpioindexl = dd->ipath_piobcnt2k; - ipath_chg_pioavailkernel(dd, 0, piobufs, 1); + /* ipath_pioavailshadow initialized earlier */ ipath_cdbg(VERBOSE, "%d PIO bufs for kernel out of %d total %u " "each for %u user ports\n", kpiobufs, piobufs, dd->ipath_pbufsport, uports); - if (dd->ipath_pioupd_thresh) { - if (dd->ipath_pbufsport < dd->ipath_pioupd_thresh) - dd->ipath_pioupd_thresh = dd->ipath_pbufsport; - if (kpiobufs < dd->ipath_pioupd_thresh) - dd->ipath_pioupd_thresh = kpiobufs; - } - ret = dd->ipath_f_early_init(dd); if (ret) { ipath_dev_err(dd, "Early initialization failure\n"); @@ -810,13 +798,6 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) } /* - * Cancel any possible active sends from early driver load. - * Follows early_init because some chips have to initialize - * PIO buffers in early_init to avoid false parity errors. - */ - ipath_cancel_sends(dd, 0); - - /* * Early_init sets rcvhdrentsize and rcvhdrsize, so this must be * done after early_init. */ @@ -836,6 +817,7 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) ipath_write_kreg(dd, dd->ipath_kregs->kr_sendpioavailaddr, dd->ipath_pioavailregs_phys); + /* * this is to detect s/w errors, which the h/w works around by * ignoring the low 6 bits of address, if it wasn't aligned. @@ -862,12 +844,6 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) ~0ULL&~INFINIPATH_HWE_MEMBISTFAILED); ipath_write_kreg(dd, dd->ipath_kregs->kr_control, 0ULL); - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - dd->ipath_sendctrl = INFINIPATH_S_PIOENABLE; - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); - /* * before error clears, since we expect serdes pll errors during * this, the first time after reset @@ -940,6 +916,19 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) else enable_chip(dd, reinit); + /* after enable_chip, so pioavailshadow setup */ + ipath_chg_pioavailkernel(dd, 0, piobufs, 1); + + /* + * Cancel any possible active sends from early driver load. + * Follows early_init because some chips have to initialize + * PIO buffers in early_init to avoid false parity errors. + * After enable and ipath_chg_pioavailkernel so we can safely + * enable pioavail updates and PIOENABLE; packets are now + * ready to go out. + */ + ipath_cancel_sends(dd, 1); + if (!reinit) { /* * Used when we close a port, for DMA already in flight diff --git a/drivers/infiniband/hw/ipath/ipath_intr.c b/drivers/infiniband/hw/ipath/ipath_intr.c index 1b58f4737c7..26900b3b7a4 100644 --- a/drivers/infiniband/hw/ipath/ipath_intr.c +++ b/drivers/infiniband/hw/ipath/ipath_intr.c @@ -38,42 +38,12 @@ #include "ipath_verbs.h" #include "ipath_common.h" -/* - * clear (write) a pio buffer, to clear a parity error. This routine - * should only be called when in freeze mode, and the buffer should be - * canceled afterwards. - */ -static void ipath_clrpiobuf(struct ipath_devdata *dd, u32 pnum) -{ - u32 __iomem *pbuf; - u32 dwcnt; /* dword count to write */ - if (pnum < dd->ipath_piobcnt2k) { - pbuf = (u32 __iomem *) (dd->ipath_pio2kbase + pnum * - dd->ipath_palign); - dwcnt = dd->ipath_piosize2k >> 2; - } - else { - pbuf = (u32 __iomem *) (dd->ipath_pio4kbase + - (pnum - dd->ipath_piobcnt2k) * dd->ipath_4kalign); - dwcnt = dd->ipath_piosize4k >> 2; - } - dev_info(&dd->pcidev->dev, - "Rewrite PIO buffer %u, to recover from parity error\n", - pnum); - - /* no flush required, since already in freeze */ - writel(dwcnt + 1, pbuf); - while (--dwcnt) - writel(0, pbuf++); -} /* * Called when we might have an error that is specific to a particular * PIO buffer, and may need to cancel that buffer, so it can be re-used. - * If rewrite is true, and bits are set in the sendbufferror registers, - * we'll write to the buffer, for error recovery on parity errors. */ -void ipath_disarm_senderrbufs(struct ipath_devdata *dd, int rewrite) +void ipath_disarm_senderrbufs(struct ipath_devdata *dd) { u32 piobcnt; unsigned long sbuf[4]; @@ -109,11 +79,8 @@ void ipath_disarm_senderrbufs(struct ipath_devdata *dd, int rewrite) } for (i = 0; i < piobcnt; i++) - if (test_bit(i, sbuf)) { - if (rewrite) - ipath_clrpiobuf(dd, i); + if (test_bit(i, sbuf)) ipath_disarm_piobufs(dd, i, 1); - } /* ignore armlaunch errs for a bit */ dd->ipath_lastcancel = jiffies+3; } @@ -164,7 +131,7 @@ static u64 handle_e_sum_errs(struct ipath_devdata *dd, ipath_err_t errs) { u64 ignore_this_time = 0; - ipath_disarm_senderrbufs(dd, 0); + ipath_disarm_senderrbufs(dd); if ((errs & E_SUM_LINK_PKTERRS) && !(dd->ipath_flags & IPATH_LINKACTIVE)) { /* @@ -909,8 +876,8 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) * processes (causing armlaunch), send errors due to going into freeze mode, * etc., and try to avoid causing extra interrupts while doing so. * Forcibly update the in-memory pioavail register copies after cleanup - * because the chip won't do it for anything changing while in freeze mode - * (we don't want to wait for the next pio buffer state change). + * because the chip won't do it while in freeze mode (the register values + * themselves are kept correct). * Make sure that we don't lose any important interrupts by using the chip * feature that says that writing 0 to a bit in *clear that is set in * *status will cause an interrupt to be generated again (if allowed by @@ -918,44 +885,23 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) */ void ipath_clear_freeze(struct ipath_devdata *dd) { - int i, im; - u64 val; - /* disable error interrupts, to avoid confusion */ ipath_write_kreg(dd, dd->ipath_kregs->kr_errormask, 0ULL); /* also disable interrupts; errormask is sometimes overwriten */ ipath_write_kreg(dd, dd->ipath_kregs->kr_intmask, 0ULL); - /* - * clear all sends, because they have may been - * completed by usercode while in freeze mode, and - * therefore would not be sent, and eventually - * might cause the process to run out of bufs - */ - ipath_cancel_sends(dd, 0); + ipath_cancel_sends(dd, 1); + + /* clear the freeze, and be sure chip saw it */ ipath_write_kreg(dd, dd->ipath_kregs->kr_control, dd->ipath_control); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - /* ensure pio avail updates continue */ + /* force in-memory update now we are out of freeze */ ipath_force_pio_avail_update(dd); /* - * We just enabled pioavailupdate, so dma copy is almost certainly - * not yet right, so read the registers directly. Similar to init - */ - for (i = 0; i < dd->ipath_pioavregs; i++) { - /* deal with 6110 chip bug */ - im = (i > 3 && (dd->ipath_flags & IPATH_SWAP_PIOBUFS)) ? - i ^ 1 : i; - val = ipath_read_kreg64(dd, (0x1000 / sizeof(u64)) + im); - dd->ipath_pioavailregs_dma[i] = cpu_to_le64(val); - dd->ipath_pioavailshadow[i] = val | - (~dd->ipath_pioavailkernel[i] << - INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT); - } - - /* * force new interrupt if any hwerr, error or interrupt bits are * still set, and clear "safe" send packet errors related to freeze * and cancelling sends. Re-enable error interrupts before possible @@ -1312,10 +1258,8 @@ irqreturn_t ipath_intr(int irq, void *data) ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); - if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA)) - handle_layer_pioavail(dd); - else - ipath_dbg("unexpected BUFAVAIL intr\n"); + /* always process; sdma verbs uses PIO for acks and VL15 */ + handle_layer_pioavail(dd); } ret = IRQ_HANDLED; diff --git a/drivers/infiniband/hw/ipath/ipath_kernel.h b/drivers/infiniband/hw/ipath/ipath_kernel.h index 202337ae90d..02b24a34059 100644 --- a/drivers/infiniband/hw/ipath/ipath_kernel.h +++ b/drivers/infiniband/hw/ipath/ipath_kernel.h @@ -117,6 +117,10 @@ struct ipath_portdata { u16 port_subport_cnt; /* non-zero if port is being shared. */ u16 port_subport_id; + /* number of pio bufs for this port (all procs, if shared) */ + u32 port_piocnt; + /* first pio buffer for this port */ + u32 port_pio_base; /* chip offset of PIO buffers for this port */ u32 port_piobufs; /* how many alloc_pages() chunks in port_rcvegrbuf_pages */ @@ -384,6 +388,8 @@ struct ipath_devdata { u32 ipath_lastrpkts; /* pio bufs allocated per port */ u32 ipath_pbufsport; + /* if remainder on bufs/port, ports < extrabuf get 1 extra */ + u32 ipath_ports_extrabuf; u32 ipath_pioupd_thresh; /* update threshold, some chips */ /* * number of ports configured as max; zero is set to number chip @@ -1011,7 +1017,7 @@ void ipath_get_eeprom_info(struct ipath_devdata *); int ipath_update_eeprom_log(struct ipath_devdata *dd); void ipath_inc_eeprom_err(struct ipath_devdata *dd, u32 eidx, u32 incr); u64 ipath_snap_cntr(struct ipath_devdata *, ipath_creg); -void ipath_disarm_senderrbufs(struct ipath_devdata *, int); +void ipath_disarm_senderrbufs(struct ipath_devdata *); void ipath_force_pio_avail_update(struct ipath_devdata *); void signal_ib_event(struct ipath_devdata *dd, enum ib_event_type ev); diff --git a/drivers/infiniband/hw/ipath/ipath_rc.c b/drivers/infiniband/hw/ipath/ipath_rc.c index c405dfba553..08b11b56761 100644 --- a/drivers/infiniband/hw/ipath/ipath_rc.c +++ b/drivers/infiniband/hw/ipath/ipath_rc.c @@ -1746,7 +1746,11 @@ void ipath_rc_rcv(struct ipath_ibdev *dev, struct ipath_ib_header *hdr, qp->r_wrid_valid = 0; wc.wr_id = qp->r_wr_id; wc.status = IB_WC_SUCCESS; - wc.opcode = IB_WC_RECV; + if (opcode == OP(RDMA_WRITE_LAST_WITH_IMMEDIATE) || + opcode == OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE)) + wc.opcode = IB_WC_RECV_RDMA_WITH_IMM; + else + wc.opcode = IB_WC_RECV; wc.vendor_err = 0; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; diff --git a/drivers/infiniband/hw/ipath/ipath_ruc.c b/drivers/infiniband/hw/ipath/ipath_ruc.c index 8ac5c1d82cc..9e3fe61cbd0 100644 --- a/drivers/infiniband/hw/ipath/ipath_ruc.c +++ b/drivers/infiniband/hw/ipath/ipath_ruc.c @@ -481,9 +481,10 @@ done: wake_up(&qp->wait); } -static void want_buffer(struct ipath_devdata *dd) +static void want_buffer(struct ipath_devdata *dd, struct ipath_qp *qp) { - if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA)) { + if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA) || + qp->ibqp.qp_type == IB_QPT_SMI) { unsigned long flags; spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); @@ -519,7 +520,7 @@ static void ipath_no_bufs_available(struct ipath_qp *qp, spin_lock_irqsave(&dev->pending_lock, flags); list_add_tail(&qp->piowait, &dev->piowait); spin_unlock_irqrestore(&dev->pending_lock, flags); - want_buffer(dev->dd); + want_buffer(dev->dd, qp); dev->n_piowait++; } diff --git a/drivers/infiniband/hw/ipath/ipath_sdma.c b/drivers/infiniband/hw/ipath/ipath_sdma.c index 1974df7a9f7..3697449c1ba 100644 --- a/drivers/infiniband/hw/ipath/ipath_sdma.c +++ b/drivers/infiniband/hw/ipath/ipath_sdma.c @@ -308,13 +308,15 @@ static void sdma_abort_task(unsigned long opaque) spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); /* - * Don't restart sdma here. Wait until link is up to ACTIVE. - * VL15 MADs used to bring the link up use PIO, and multiple - * link transitions otherwise cause the sdma engine to be + * Don't restart sdma here (with the exception + * below). Wait until link is up to ACTIVE. VL15 MADs + * used to bring the link up use PIO, and multiple link + * transitions otherwise cause the sdma engine to be * stopped and started multiple times. - * The disable is done here, including the shadow, so the - * state is kept consistent. - * See ipath_restart_sdma() for the actual starting of sdma. + * The disable is done here, including the shadow, + * so the state is kept consistent. + * See ipath_restart_sdma() for the actual starting + * of sdma. */ spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); dd->ipath_sendctrl &= ~INFINIPATH_S_SDMAENABLE; @@ -326,6 +328,13 @@ static void sdma_abort_task(unsigned long opaque) /* make sure I see next message */ dd->ipath_sdma_abort_jiffies = 0; + /* + * Not everything that takes SDMA offline is a link + * status change. If the link was up, restart SDMA. + */ + if (dd->ipath_flags & IPATH_LINKACTIVE) + ipath_restart_sdma(dd); + goto done; } @@ -427,7 +436,12 @@ int setup_sdma(struct ipath_devdata *dd) goto done; } - dd->ipath_sdma_status = 0; + /* + * Set initial status as if we had been up, then gone down. + * This lets initial start on transition to ACTIVE be the + * same as restart after link flap. + */ + dd->ipath_sdma_status = IPATH_SDMA_ABORT_ABORTED; dd->ipath_sdma_abort_jiffies = 0; dd->ipath_sdma_generation = 0; dd->ipath_sdma_descq_tail = 0; @@ -449,16 +463,19 @@ int setup_sdma(struct ipath_devdata *dd) ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmaheadaddr, dd->ipath_sdma_head_phys); - /* Reserve all the former "kernel" piobufs */ - n = dd->ipath_piobcnt2k + dd->ipath_piobcnt4k - dd->ipath_pioreserved; - for (i = dd->ipath_lastport_piobuf; i < n; ++i) { + /* + * Reserve all the former "kernel" piobufs, using high number range + * so we get as many 4K buffers as possible + */ + n = dd->ipath_piobcnt2k + dd->ipath_piobcnt4k; + i = dd->ipath_lastport_piobuf + dd->ipath_pioreserved; + ipath_chg_pioavailkernel(dd, i, n - i , 0); + for (; i < n; ++i) { unsigned word = i / 64; unsigned bit = i & 63; BUG_ON(word >= 3); senddmabufmask[word] |= 1ULL << bit; } - ipath_chg_pioavailkernel(dd, dd->ipath_lastport_piobuf, - n - dd->ipath_lastport_piobuf, 0); ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask0, senddmabufmask[0]); ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask1, @@ -615,6 +632,9 @@ void ipath_restart_sdma(struct ipath_devdata *dd) ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + /* notify upper layers */ + ipath_ib_piobufavail(dd->verbs_dev); + bail: return; } diff --git a/drivers/infiniband/hw/ipath/ipath_verbs.c b/drivers/infiniband/hw/ipath/ipath_verbs.c index e63927cce5b..5015cd2e57b 100644 --- a/drivers/infiniband/hw/ipath/ipath_verbs.c +++ b/drivers/infiniband/hw/ipath/ipath_verbs.c @@ -396,7 +396,6 @@ static int ipath_post_one_send(struct ipath_qp *qp, struct ib_send_wr *wr) wqe = get_swqe_ptr(qp, qp->s_head); wqe->wr = *wr; - wqe->ssn = qp->s_ssn++; wqe->length = 0; if (wr->num_sge) { acc = wr->opcode >= IB_WR_RDMA_READ ? @@ -422,6 +421,7 @@ static int ipath_post_one_send(struct ipath_qp *qp, struct ib_send_wr *wr) goto bail_inval; } else if (wqe->length > to_idev(qp->ibqp.device)->dd->ipath_ibmtu) goto bail_inval; + wqe->ssn = qp->s_ssn++; qp->s_head = next; ret = 0; diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 5938fa96292..faf3d891297 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -886,7 +886,7 @@ static int make_request(struct request_queue *q, struct bio * bio) */ raid10_find_phys(conf, r10_bio); retry_write: - blocked_rdev = 0; + blocked_rdev = NULL; rcu_read_lock(); for (i = 0; i < conf->copies; i++) { int d = r10_bio->devs[i].devnum; diff --git a/drivers/media/Makefile b/drivers/media/Makefile index 73f742c7e81..cc11c4c0e7e 100644 --- a/drivers/media/Makefile +++ b/drivers/media/Makefile @@ -2,6 +2,8 @@ # Makefile for the kernel multimedia device drivers. # +obj-y := common/ + obj-$(CONFIG_VIDEO_MEDIA) += common/ # Since hybrid devices are here, should be compiled if DVB and/or V4L diff --git a/drivers/media/video/cx18/cx18-driver.c b/drivers/media/video/cx18/cx18-driver.c index 8f5ed9b4bf8..3f55d47bc4b 100644 --- a/drivers/media/video/cx18/cx18-driver.c +++ b/drivers/media/video/cx18/cx18-driver.c @@ -613,7 +613,7 @@ static int __devinit cx18_probe(struct pci_dev *dev, } cx = kzalloc(sizeof(struct cx18), GFP_ATOMIC); - if (cx == 0) { + if (!cx) { spin_unlock(&cx18_cards_lock); return -ENOMEM; } diff --git a/drivers/media/video/saa7134/saa7134-video.c b/drivers/media/video/saa7134/saa7134-video.c index a0baf2d0ba7..48e1a01718e 100644 --- a/drivers/media/video/saa7134/saa7134-video.c +++ b/drivers/media/video/saa7134/saa7134-video.c @@ -1634,7 +1634,7 @@ static int saa7134_s_fmt_overlay(struct file *file, void *priv, struct saa7134_fh *fh = priv; struct saa7134_dev *dev = fh->dev; int err; - unsigned int flags; + unsigned long flags; if (saa7134_no_overlay > 0) { printk(KERN_ERR "V4L2_BUF_TYPE_VIDEO_OVERLAY: no_overlay\n"); diff --git a/drivers/net/irda/nsc-ircc.c b/drivers/net/irda/nsc-ircc.c index a873d2b315c..a7714da7c28 100644 --- a/drivers/net/irda/nsc-ircc.c +++ b/drivers/net/irda/nsc-ircc.c @@ -100,7 +100,9 @@ static int nsc_ircc_probe_39x(nsc_chip_t *chip, chipio_t *info); static int nsc_ircc_init_108(nsc_chip_t *chip, chipio_t *info); static int nsc_ircc_init_338(nsc_chip_t *chip, chipio_t *info); static int nsc_ircc_init_39x(nsc_chip_t *chip, chipio_t *info); +#ifdef CONFIG_PNP static int nsc_ircc_pnp_probe(struct pnp_dev *dev, const struct pnp_device_id *id); +#endif /* These are the known NSC chips */ static nsc_chip_t chips[] = { @@ -156,9 +158,11 @@ static const struct pnp_device_id nsc_ircc_pnp_table[] = { MODULE_DEVICE_TABLE(pnp, nsc_ircc_pnp_table); static struct pnp_driver nsc_ircc_pnp_driver = { +#ifdef CONFIG_PNP .name = "nsc-ircc", .id_table = nsc_ircc_pnp_table, .probe = nsc_ircc_pnp_probe, +#endif }; /* Some prototypes */ @@ -916,6 +920,7 @@ static int nsc_ircc_probe_39x(nsc_chip_t *chip, chipio_t *info) return 0; } +#ifdef CONFIG_PNP /* PNP probing */ static int nsc_ircc_pnp_probe(struct pnp_dev *dev, const struct pnp_device_id *id) { @@ -952,6 +957,7 @@ static int nsc_ircc_pnp_probe(struct pnp_dev *dev, const struct pnp_device_id *i return 0; } +#endif /* * Function nsc_ircc_setup (info) diff --git a/drivers/net/irda/smsc-ircc2.c b/drivers/net/irda/smsc-ircc2.c index 1f26da761e9..cfe0194fef7 100644 --- a/drivers/net/irda/smsc-ircc2.c +++ b/drivers/net/irda/smsc-ircc2.c @@ -376,6 +376,7 @@ MODULE_DEVICE_TABLE(pnp, smsc_ircc_pnp_table); static int pnp_driver_registered; +#ifdef CONFIG_PNP static int __init smsc_ircc_pnp_probe(struct pnp_dev *dev, const struct pnp_device_id *dev_id) { @@ -402,7 +403,9 @@ static struct pnp_driver smsc_ircc_pnp_driver = { .id_table = smsc_ircc_pnp_table, .probe = smsc_ircc_pnp_probe, }; - +#else /* CONFIG_PNP */ +static struct pnp_driver smsc_ircc_pnp_driver; +#endif /******************************************************************************* * diff --git a/drivers/net/niu.c b/drivers/net/niu.c index 4009c4ce96b..57cfd72ffdf 100644 --- a/drivers/net/niu.c +++ b/drivers/net/niu.c @@ -1,6 +1,6 @@ /* niu.c: Neptune ethernet driver. * - * Copyright (C) 2007 David S. Miller (davem@davemloft.net) + * Copyright (C) 2007, 2008 David S. Miller (davem@davemloft.net) */ #include <linux/module.h> @@ -33,8 +33,8 @@ #define DRV_MODULE_NAME "niu" #define PFX DRV_MODULE_NAME ": " -#define DRV_MODULE_VERSION "0.8" -#define DRV_MODULE_RELDATE "April 24, 2008" +#define DRV_MODULE_VERSION "0.9" +#define DRV_MODULE_RELDATE "May 4, 2008" static char version[] __devinitdata = DRV_MODULE_NAME ".c:v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n"; @@ -7264,8 +7264,11 @@ static int __devinit niu_get_and_validate_port(struct niu *np) parent->num_ports = nr64(ESPC_NUM_PORTS_MACS) & ESPC_NUM_PORTS_MACS_VAL; + /* All of the current probing methods fail on + * Maramba on-board parts. + */ if (!parent->num_ports) - return -ENODEV; + parent->num_ports = 4; } } } diff --git a/drivers/net/usb/asix.c b/drivers/net/usb/asix.c index 6f245cfb662..dc6f097062d 100644 --- a/drivers/net/usb/asix.c +++ b/drivers/net/usb/asix.c @@ -1381,6 +1381,10 @@ static const struct usb_device_id products [] = { USB_DEVICE (0x0411, 0x003d), .driver_info = (unsigned long) &ax8817x_info, }, { + // Buffalo LUA-U2-GT 10/100/1000 + USB_DEVICE (0x0411, 0x006e), + .driver_info = (unsigned long) &ax88178_info, +}, { // Sitecom LN-029 "USB 2.0 10/100 Ethernet adapter" USB_DEVICE (0x6189, 0x182d), .driver_info = (unsigned long) &ax8817x_info, diff --git a/drivers/net/wan/lapbether.c b/drivers/net/wan/lapbether.c index b5860b97a93..24fd613466b 100644 --- a/drivers/net/wan/lapbether.c +++ b/drivers/net/wan/lapbether.c @@ -459,6 +459,7 @@ static void __exit lapbeth_cleanup_driver(void) list_for_each_safe(entry, tmp, &lapbeth_devices) { lapbeth = list_entry(entry, struct lapbethdev, node); + dev_put(lapbeth->ethdev); unregister_netdevice(lapbeth->axdev); } rtnl_unlock(); diff --git a/drivers/net/wireless/iwlwifi/Kconfig b/drivers/net/wireless/iwlwifi/Kconfig index d5b7a76fcaa..62fb89d8231 100644 --- a/drivers/net/wireless/iwlwifi/Kconfig +++ b/drivers/net/wireless/iwlwifi/Kconfig @@ -1,6 +1,5 @@ config IWLWIFI - bool - default n + tristate config IWLCORE tristate "Intel Wireless Wifi Core" diff --git a/drivers/rtc/rtc-ds1511.c b/drivers/rtc/rtc-ds1511.c index a83a40b3eba..0f0d27d1c4c 100644 --- a/drivers/rtc/rtc-ds1511.c +++ b/drivers/rtc/rtc-ds1511.c @@ -184,7 +184,7 @@ ds1511_wdog_disable(void) static int ds1511_rtc_set_time(struct device *dev, struct rtc_time *rtc_tm) { u8 mon, day, dow, hrs, min, sec, yrs, cen; - unsigned int flags; + unsigned long flags; /* * won't have to change this for a while @@ -247,7 +247,7 @@ static int ds1511_rtc_set_time(struct device *dev, struct rtc_time *rtc_tm) static int ds1511_rtc_read_time(struct device *dev, struct rtc_time *rtc_tm) { unsigned int century; - unsigned int flags; + unsigned long flags; spin_lock_irqsave(&ds1511_lock, flags); rtc_disable_update(); diff --git a/drivers/s390/char/tty3270.c b/drivers/s390/char/tty3270.c index c1f2adefad4..5043150019a 100644 --- a/drivers/s390/char/tty3270.c +++ b/drivers/s390/char/tty3270.c @@ -965,8 +965,7 @@ tty3270_write_room(struct tty_struct *tty) * Insert character into the screen at the current position with the * current color and highlight. This function does NOT do cursor movement. */ -static int -tty3270_put_character(struct tty3270 *tp, char ch) +static void tty3270_put_character(struct tty3270 *tp, char ch) { struct tty3270_line *line; struct tty3270_cell *cell; @@ -986,7 +985,6 @@ tty3270_put_character(struct tty3270 *tp, char ch) cell->character = tp->view.ascebc[(unsigned int) ch]; cell->highlight = tp->highlight; cell->f_color = tp->f_color; - return 1; } /* @@ -1612,16 +1610,15 @@ tty3270_write(struct tty_struct * tty, /* * Put single characters to the ttys character buffer */ -static void -tty3270_put_char(struct tty_struct *tty, unsigned char ch) +static int tty3270_put_char(struct tty_struct *tty, unsigned char ch) { struct tty3270 *tp; tp = tty->driver_data; - if (!tp) - return; - if (tp->char_count < TTY3270_CHAR_BUF_SIZE) - tp->char_buf[tp->char_count++] = ch; + if (!tp || tp->char_count >= TTY3270_CHAR_BUF_SIZE) + return 0; + tp->char_buf[tp->char_count++] = ch; + return 1; } /* diff --git a/drivers/s390/cio/blacklist.c b/drivers/s390/cio/blacklist.c index 40ef948fcb3..9c21b8f43f9 100644 --- a/drivers/s390/cio/blacklist.c +++ b/drivers/s390/cio/blacklist.c @@ -19,6 +19,7 @@ #include <asm/cio.h> #include <asm/uaccess.h> +#include <asm/cio.h> #include "blacklist.h" #include "cio.h" @@ -43,164 +44,169 @@ typedef enum {add, free} range_action; * Function: blacklist_range * (Un-)blacklist the devices from-to */ -static void -blacklist_range (range_action action, unsigned int from, unsigned int to, - unsigned int ssid) +static int blacklist_range(range_action action, unsigned int from_ssid, + unsigned int to_ssid, unsigned int from, + unsigned int to, int msgtrigger) { - if (!to) - to = from; - - if (from > to || to > __MAX_SUBCHANNEL || ssid > __MAX_SSID) { - printk (KERN_WARNING "cio: Invalid blacklist range " - "0.%x.%04x to 0.%x.%04x, skipping\n", - ssid, from, ssid, to); - return; + if ((from_ssid > to_ssid) || ((from_ssid == to_ssid) && (from > to))) { + if (msgtrigger) + printk(KERN_WARNING "cio: Invalid cio_ignore range " + "0.%x.%04x-0.%x.%04x\n", from_ssid, from, + to_ssid, to); + return 1; } - for (; from <= to; from++) { + + while ((from_ssid < to_ssid) || ((from_ssid == to_ssid) && + (from <= to))) { if (action == add) - set_bit (from, bl_dev[ssid]); + set_bit(from, bl_dev[from_ssid]); else - clear_bit (from, bl_dev[ssid]); + clear_bit(from, bl_dev[from_ssid]); + from++; + if (from > __MAX_SUBCHANNEL) { + from_ssid++; + from = 0; + } } + + return 0; } -/* - * Function: blacklist_busid - * Get devno/busid from given string. - * Shamelessly grabbed from dasd_devmap.c. - */ -static int -blacklist_busid(char **str, int *id0, int *ssid, int *devno) +static int pure_hex(char **cp, unsigned int *val, int min_digit, + int max_digit, int max_val) { - int val, old_style; - char *sav; + int diff; + unsigned int value; - sav = *str; + diff = 0; + *val = 0; - /* check for leading '0x' */ - old_style = 0; - if ((*str)[0] == '0' && (*str)[1] == 'x') { - *str += 2; - old_style = 1; - } - if (!isxdigit((*str)[0])) /* We require at least one hex digit */ - goto confused; - val = simple_strtoul(*str, str, 16); - if (old_style || (*str)[0] != '.') { - *id0 = *ssid = 0; - if (val < 0 || val > 0xffff) - goto confused; - *devno = val; - if ((*str)[0] != ',' && (*str)[0] != '-' && - (*str)[0] != '\n' && (*str)[0] != '\0') - goto confused; - return 0; + while (isxdigit(**cp) && (diff <= max_digit)) { + + if (isdigit(**cp)) + value = **cp - '0'; + else + value = tolower(**cp) - 'a' + 10; + *val = *val * 16 + value; + (*cp)++; + diff++; } - /* New style x.y.z busid */ - if (val < 0 || val > 0xff) - goto confused; - *id0 = val; - (*str)++; - if (!isxdigit((*str)[0])) /* We require at least one hex digit */ - goto confused; - val = simple_strtoul(*str, str, 16); - if (val < 0 || val > 0xff || (*str)++[0] != '.') - goto confused; - *ssid = val; - if (!isxdigit((*str)[0])) /* We require at least one hex digit */ - goto confused; - val = simple_strtoul(*str, str, 16); - if (val < 0 || val > 0xffff) - goto confused; - *devno = val; - if ((*str)[0] != ',' && (*str)[0] != '-' && - (*str)[0] != '\n' && (*str)[0] != '\0') - goto confused; + + if ((diff < min_digit) || (diff > max_digit) || (*val > max_val)) + return 1; + return 0; -confused: - strsep(str, ",\n"); - printk(KERN_WARNING "cio: Invalid cio_ignore parameter '%s'\n", sav); - return 1; } -static int -blacklist_parse_parameters (char *str, range_action action) +static int parse_busid(char *str, int *cssid, int *ssid, int *devno, + int msgtrigger) { - int from, to, from_id0, to_id0, from_ssid, to_ssid; - - while (*str != 0 && *str != '\n') { - range_action ra = action; - while(*str == ',') - str++; - if (*str == '!') { - ra = !action; - ++str; + char *str_work; + int val, rc, ret; + + rc = 1; + + if (*str == '\0') + goto out; + + /* old style */ + str_work = str; + val = simple_strtoul(str, &str_work, 16); + + if (*str_work == '\0') { + if (val <= __MAX_SUBCHANNEL) { + *devno = val; + *ssid = 0; + *cssid = 0; + rc = 0; } + goto out; + } - /* - * Since we have to parse the proc commands and the - * kernel arguments we have to check four cases - */ - if (strncmp(str,"all,",4) == 0 || strcmp(str,"all") == 0 || - strncmp(str,"all\n",4) == 0 || strncmp(str,"all ",4) == 0) { - int j; - - str += 3; - for (j=0; j <= __MAX_SSID; j++) - blacklist_range(ra, 0, __MAX_SUBCHANNEL, j); - } else { - int rc; + /* new style */ + str_work = str; + ret = pure_hex(&str_work, cssid, 1, 2, __MAX_CSSID); + if (ret || (str_work[0] != '.')) + goto out; + str_work++; + ret = pure_hex(&str_work, ssid, 1, 1, __MAX_SSID); + if (ret || (str_work[0] != '.')) + goto out; + str_work++; + ret = pure_hex(&str_work, devno, 4, 4, __MAX_SUBCHANNEL); + if (ret || (str_work[0] != '\0')) + goto out; + + rc = 0; +out: + if (rc && msgtrigger) + printk(KERN_WARNING "cio: Invalid cio_ignore device '%s'\n", + str); + + return rc; +} - rc = blacklist_busid(&str, &from_id0, - &from_ssid, &from); - if (rc) - continue; - to = from; - to_id0 = from_id0; - to_ssid = from_ssid; - if (*str == '-') { - str++; - rc = blacklist_busid(&str, &to_id0, - &to_ssid, &to); - if (rc) - continue; - } - if (*str == '-') { - printk(KERN_WARNING "cio: invalid cio_ignore " - "parameter '%s'\n", - strsep(&str, ",\n")); - continue; - } - if ((from_id0 != to_id0) || - (from_ssid != to_ssid)) { - printk(KERN_WARNING "cio: invalid cio_ignore " - "range %x.%x.%04x-%x.%x.%04x\n", - from_id0, from_ssid, from, - to_id0, to_ssid, to); - continue; +static int blacklist_parse_parameters(char *str, range_action action, + int msgtrigger) +{ + int from_cssid, to_cssid, from_ssid, to_ssid, from, to; + int rc, totalrc; + char *parm; + range_action ra; + + totalrc = 0; + + while ((parm = strsep(&str, ","))) { + rc = 0; + ra = action; + if (*parm == '!') { + if (ra == add) + ra = free; + else + ra = add; + parm++; + } + if (strcmp(parm, "all") == 0) { + from_cssid = 0; + from_ssid = 0; + from = 0; + to_cssid = __MAX_CSSID; + to_ssid = __MAX_SSID; + to = __MAX_SUBCHANNEL; + } else { + rc = parse_busid(strsep(&parm, "-"), &from_cssid, + &from_ssid, &from, msgtrigger); + if (!rc) { + if (parm != NULL) + rc = parse_busid(parm, &to_cssid, + &to_ssid, &to, + msgtrigger); + else { + to_cssid = from_cssid; + to_ssid = from_ssid; + to = from; + } } - blacklist_range (ra, from, to, to_ssid); } + if (!rc) { + rc = blacklist_range(ra, from_ssid, to_ssid, from, to, + msgtrigger); + if (rc) + totalrc = 1; + } else + totalrc = 1; } - return 1; + + return totalrc; } -/* Parsing the commandline for blacklist parameters, e.g. to blacklist - * bus ids 0.0.1234, 0.0.1235 and 0.0.1236, you could use any of: - * - cio_ignore=1234-1236 - * - cio_ignore=0x1234-0x1235,1236 - * - cio_ignore=0x1234,1235-1236 - * - cio_ignore=1236 cio_ignore=1234-0x1236 - * - cio_ignore=1234 cio_ignore=1236 cio_ignore=0x1235 - * - cio_ignore=0.0.1234-0.0.1236 - * - cio_ignore=0.0.1234,0x1235,1236 - * - ... - */ static int __init blacklist_setup (char *str) { CIO_MSG_EVENT(6, "Reading blacklist parameters\n"); - return blacklist_parse_parameters (str, add); + if (blacklist_parse_parameters(str, add, 1)) + return 0; + return 1; } __setup ("cio_ignore=", blacklist_setup); @@ -224,27 +230,23 @@ is_blacklisted (int ssid, int devno) * Function: blacklist_parse_proc_parameters * parse the stuff which is piped to /proc/cio_ignore */ -static void -blacklist_parse_proc_parameters (char *buf) +static int blacklist_parse_proc_parameters(char *buf) { - if (strncmp (buf, "free ", 5) == 0) { - blacklist_parse_parameters (buf + 5, free); - } else if (strncmp (buf, "add ", 4) == 0) { - /* - * We don't need to check for known devices since - * css_probe_device will handle this correctly. - */ - blacklist_parse_parameters (buf + 4, add); - } else { - printk (KERN_WARNING "cio: cio_ignore: Parse error; \n" - KERN_WARNING "try using 'free all|<devno-range>," - "<devno-range>,...'\n" - KERN_WARNING "or 'add <devno-range>," - "<devno-range>,...'\n"); - return; - } + int rc; + char *parm; + + parm = strsep(&buf, " "); + + if (strcmp("free", parm) == 0) + rc = blacklist_parse_parameters(buf, free, 0); + else if (strcmp("add", parm) == 0) + rc = blacklist_parse_parameters(buf, add, 0); + else + return 1; css_schedule_reprobe(); + + return rc; } /* Iterator struct for all devices. */ @@ -328,6 +330,8 @@ cio_ignore_write(struct file *file, const char __user *user_buf, size_t user_len, loff_t *offset) { char *buf; + size_t i; + ssize_t rc, ret; if (*offset) return -EINVAL; @@ -336,16 +340,27 @@ cio_ignore_write(struct file *file, const char __user *user_buf, buf = vmalloc (user_len + 1); /* maybe better use the stack? */ if (buf == NULL) return -ENOMEM; + memset(buf, 0, user_len + 1); + if (strncpy_from_user (buf, user_buf, user_len) < 0) { - vfree (buf); - return -EFAULT; + rc = -EFAULT; + goto out_free; } - buf[user_len] = '\0'; - blacklist_parse_proc_parameters (buf); + i = user_len - 1; + while ((i >= 0) && (isspace(buf[i]) || (buf[i] == 0))) { + buf[i] = '\0'; + i--; + } + ret = blacklist_parse_proc_parameters(buf); + if (ret) + rc = -EINVAL; + else + rc = user_len; +out_free: vfree (buf); - return user_len; + return rc; } static const struct seq_operations cio_ignore_proc_seq_ops = { diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index 08a57816130..82c6a2d4512 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -39,23 +39,6 @@ debug_info_t *cio_debug_msg_id; debug_info_t *cio_debug_trace_id; debug_info_t *cio_debug_crw_id; -int cio_show_msg; - -static int __init -cio_setup (char *parm) -{ - if (!strcmp (parm, "yes")) - cio_show_msg = 1; - else if (!strcmp (parm, "no")) - cio_show_msg = 0; - else - printk(KERN_ERR "cio: cio_setup: " - "invalid cio_msg parameter '%s'", parm); - return 1; -} - -__setup ("cio_msg=", cio_setup); - /* * Function: cio_debug_init * Initializes three debug logs for common I/O: @@ -166,7 +149,7 @@ cio_start_handle_notoper(struct subchannel *sch, __u8 lpm) stsch (sch->schid, &sch->schib); - CIO_MSG_EVENT(0, "cio_start: 'not oper' status for " + CIO_MSG_EVENT(2, "cio_start: 'not oper' status for " "subchannel 0.%x.%04x!\n", sch->schid.ssid, sch->schid.sch_no); sprintf(dbf_text, "no%s", sch->dev.bus_id); @@ -567,10 +550,9 @@ cio_validate_subchannel (struct subchannel *sch, struct subchannel_id schid) * ... just being curious we check for non I/O subchannels */ if (sch->st != 0) { - CIO_DEBUG(KERN_INFO, 0, - "Subchannel 0.%x.%04x reports " - "non-I/O subchannel type %04X\n", - sch->schid.ssid, sch->schid.sch_no, sch->st); + CIO_MSG_EVENT(4, "Subchannel 0.%x.%04x reports " + "non-I/O subchannel type %04X\n", + sch->schid.ssid, sch->schid.sch_no, sch->st); /* We stop here for non-io subchannels. */ err = sch->st; goto out; @@ -588,7 +570,7 @@ cio_validate_subchannel (struct subchannel *sch, struct subchannel_id schid) * This device must not be known to Linux. So we simply * say that there is no device and return ENODEV. */ - CIO_MSG_EVENT(4, "Blacklisted device detected " + CIO_MSG_EVENT(6, "Blacklisted device detected " "at devno %04X, subchannel set %x\n", sch->schib.pmcw.dev, sch->schid.ssid); err = -ENODEV; @@ -601,12 +583,11 @@ cio_validate_subchannel (struct subchannel *sch, struct subchannel_id schid) sch->lpm = sch->schib.pmcw.pam & sch->opm; sch->isc = 3; - CIO_DEBUG(KERN_INFO, 0, - "Detected device %04x on subchannel 0.%x.%04X" - " - PIM = %02X, PAM = %02X, POM = %02X\n", - sch->schib.pmcw.dev, sch->schid.ssid, - sch->schid.sch_no, sch->schib.pmcw.pim, - sch->schib.pmcw.pam, sch->schib.pmcw.pom); + CIO_MSG_EVENT(6, "Detected device %04x on subchannel 0.%x.%04X " + "- PIM = %02X, PAM = %02X, POM = %02X\n", + sch->schib.pmcw.dev, sch->schid.ssid, + sch->schid.sch_no, sch->schib.pmcw.pim, + sch->schib.pmcw.pam, sch->schib.pmcw.pom); /* * We now have to initially ... diff --git a/drivers/s390/cio/cio.h b/drivers/s390/cio/cio.h index 3c75412904d..6e933aebe01 100644 --- a/drivers/s390/cio/cio.h +++ b/drivers/s390/cio/cio.h @@ -118,6 +118,4 @@ extern void *cio_get_console_priv(void); #define cio_get_console_priv() NULL #endif -extern int cio_show_msg; - #endif diff --git a/drivers/s390/cio/cio_debug.h b/drivers/s390/cio/cio_debug.h index d7429ef6c66..e64e8278c42 100644 --- a/drivers/s390/cio/cio_debug.h +++ b/drivers/s390/cio/cio_debug.h @@ -31,10 +31,4 @@ static inline void CIO_HEX_EVENT(int level, void *data, int length) } } -#define CIO_DEBUG(printk_level, event_level, msg...) do { \ - if (cio_show_msg) \ - printk(printk_level "cio: " msg); \ - CIO_MSG_EVENT(event_level, msg); \ - } while (0) - #endif diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c index 595e327d2f7..a76956512b2 100644 --- a/drivers/s390/cio/css.c +++ b/drivers/s390/cio/css.c @@ -570,7 +570,7 @@ static void reprobe_all(struct work_struct *unused) { int ret; - CIO_MSG_EVENT(2, "reprobe start\n"); + CIO_MSG_EVENT(4, "reprobe start\n"); need_reprobe = 0; /* Make sure initial subchannel scan is done. */ @@ -578,7 +578,7 @@ static void reprobe_all(struct work_struct *unused) atomic_read(&ccw_device_init_count) == 0); ret = for_each_subchannel_staged(NULL, reprobe_subchannel, NULL); - CIO_MSG_EVENT(2, "reprobe done (rc=%d, need_reprobe=%d)\n", ret, + CIO_MSG_EVENT(4, "reprobe done (rc=%d, need_reprobe=%d)\n", ret, need_reprobe); } diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index abfd601d237..e22813db74a 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -341,7 +341,7 @@ ccw_device_remove_disconnected(struct ccw_device *cdev) rc = device_schedule_callback(&cdev->dev, ccw_device_remove_orphan_cb); if (rc) - CIO_MSG_EVENT(2, "Couldn't unregister orphan " + CIO_MSG_EVENT(0, "Couldn't unregister orphan " "0.%x.%04x\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -351,7 +351,7 @@ ccw_device_remove_disconnected(struct ccw_device *cdev) rc = device_schedule_callback(cdev->dev.parent, ccw_device_remove_sch_cb); if (rc) - CIO_MSG_EVENT(2, "Couldn't unregister disconnected device " + CIO_MSG_EVENT(0, "Couldn't unregister disconnected device " "0.%x.%04x\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -397,7 +397,7 @@ int ccw_device_set_offline(struct ccw_device *cdev) if (ret == 0) wait_event(cdev->private->wait_q, dev_fsm_final_state(cdev)); else { - CIO_MSG_EVENT(2, "ccw_device_offline returned %d, " + CIO_MSG_EVENT(0, "ccw_device_offline returned %d, " "device 0.%x.%04x\n", ret, cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -433,7 +433,7 @@ int ccw_device_set_online(struct ccw_device *cdev) if (ret == 0) wait_event(cdev->private->wait_q, dev_fsm_final_state(cdev)); else { - CIO_MSG_EVENT(2, "ccw_device_online returned %d, " + CIO_MSG_EVENT(0, "ccw_device_online returned %d, " "device 0.%x.%04x\n", ret, cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -451,7 +451,7 @@ int ccw_device_set_online(struct ccw_device *cdev) if (ret == 0) wait_event(cdev->private->wait_q, dev_fsm_final_state(cdev)); else - CIO_MSG_EVENT(2, "ccw_device_offline returned %d, " + CIO_MSG_EVENT(0, "ccw_device_offline returned %d, " "device 0.%x.%04x\n", ret, cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -803,7 +803,7 @@ static void sch_attach_disconnected_device(struct subchannel *sch, other_sch = to_subchannel(get_device(cdev->dev.parent)); ret = device_move(&cdev->dev, &sch->dev); if (ret) { - CIO_MSG_EVENT(2, "Moving disconnected device 0.%x.%04x failed " + CIO_MSG_EVENT(0, "Moving disconnected device 0.%x.%04x failed " "(ret=%d)!\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno, ret); put_device(&other_sch->dev); @@ -933,7 +933,7 @@ io_subchannel_register(struct work_struct *work) ret = device_reprobe(&cdev->dev); if (ret) /* We can't do much here. */ - CIO_MSG_EVENT(2, "device_reprobe() returned" + CIO_MSG_EVENT(0, "device_reprobe() returned" " %d for 0.%x.%04x\n", ret, cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -1086,7 +1086,7 @@ static void ccw_device_move_to_sch(struct work_struct *work) rc = device_move(&cdev->dev, &sch->dev); mutex_unlock(&sch->reg_mutex); if (rc) { - CIO_MSG_EVENT(2, "Moving device 0.%x.%04x to subchannel " + CIO_MSG_EVENT(0, "Moving device 0.%x.%04x to subchannel " "0.%x.%04x failed (ret=%d)!\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno, sch->schid.ssid, @@ -1446,8 +1446,7 @@ ccw_device_remove (struct device *dev) wait_event(cdev->private->wait_q, dev_fsm_final_state(cdev)); else - //FIXME: we can't fail! - CIO_MSG_EVENT(2, "ccw_device_offline returned %d, " + CIO_MSG_EVENT(0, "ccw_device_offline returned %d, " "device 0.%x.%04x\n", ret, cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -1524,7 +1523,7 @@ static int recovery_check(struct device *dev, void *data) spin_lock_irq(cdev->ccwlock); switch (cdev->private->state) { case DEV_STATE_DISCONNECTED: - CIO_MSG_EVENT(3, "recovery: trigger 0.%x.%04x\n", + CIO_MSG_EVENT(4, "recovery: trigger 0.%x.%04x\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno); dev_fsm_event(cdev, DEV_EVENT_VERIFY); @@ -1554,7 +1553,7 @@ static void recovery_work_func(struct work_struct *unused) } spin_unlock_irq(&recovery_lock); } else - CIO_MSG_EVENT(2, "recovery: end\n"); + CIO_MSG_EVENT(4, "recovery: end\n"); } static DECLARE_WORK(recovery_work, recovery_work_func); @@ -1572,7 +1571,7 @@ void ccw_device_schedule_recovery(void) { unsigned long flags; - CIO_MSG_EVENT(2, "recovery: schedule\n"); + CIO_MSG_EVENT(4, "recovery: schedule\n"); spin_lock_irqsave(&recovery_lock, flags); if (!timer_pending(&recovery_timer) || (recovery_phase != 0)) { recovery_phase = 0; diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c index 99403b0a97a..e268d5a77c1 100644 --- a/drivers/s390/cio/device_fsm.c +++ b/drivers/s390/cio/device_fsm.c @@ -322,10 +322,10 @@ ccw_device_recog_done(struct ccw_device *cdev, int state) same_dev = 0; /* Keep the compiler quiet... */ switch (state) { case DEV_STATE_NOT_OPER: - CIO_DEBUG(KERN_WARNING, 2, - "SenseID : unknown device %04x on subchannel " - "0.%x.%04x\n", cdev->private->dev_id.devno, - sch->schid.ssid, sch->schid.sch_no); + CIO_MSG_EVENT(2, "SenseID : unknown device %04x on " + "subchannel 0.%x.%04x\n", + cdev->private->dev_id.devno, + sch->schid.ssid, sch->schid.sch_no); break; case DEV_STATE_OFFLINE: if (cdev->private->state == DEV_STATE_DISCONNECTED_SENSE_ID) { @@ -348,20 +348,19 @@ ccw_device_recog_done(struct ccw_device *cdev, int state) return; } /* Issue device info message. */ - CIO_DEBUG(KERN_INFO, 2, - "SenseID : device 0.%x.%04x reports: " - "CU Type/Mod = %04X/%02X, Dev Type/Mod = " - "%04X/%02X\n", - cdev->private->dev_id.ssid, - cdev->private->dev_id.devno, - cdev->id.cu_type, cdev->id.cu_model, - cdev->id.dev_type, cdev->id.dev_model); + CIO_MSG_EVENT(4, "SenseID : device 0.%x.%04x reports: " + "CU Type/Mod = %04X/%02X, Dev Type/Mod = " + "%04X/%02X\n", + cdev->private->dev_id.ssid, + cdev->private->dev_id.devno, + cdev->id.cu_type, cdev->id.cu_model, + cdev->id.dev_type, cdev->id.dev_model); break; case DEV_STATE_BOXED: - CIO_DEBUG(KERN_WARNING, 2, - "SenseID : boxed device %04x on subchannel " - "0.%x.%04x\n", cdev->private->dev_id.devno, - sch->schid.ssid, sch->schid.sch_no); + CIO_MSG_EVENT(0, "SenseID : boxed device %04x on " + " subchannel 0.%x.%04x\n", + cdev->private->dev_id.devno, + sch->schid.ssid, sch->schid.sch_no); break; } cdev->private->state = state; @@ -443,9 +442,8 @@ ccw_device_done(struct ccw_device *cdev, int state) if (state == DEV_STATE_BOXED) - CIO_DEBUG(KERN_WARNING, 2, - "Boxed device %04x on subchannel %04x\n", - cdev->private->dev_id.devno, sch->schid.sch_no); + CIO_MSG_EVENT(0, "Boxed device %04x on subchannel %04x\n", + cdev->private->dev_id.devno, sch->schid.sch_no); if (cdev->private->flags.donotify) { cdev->private->flags.donotify = 0; @@ -900,7 +898,7 @@ ccw_device_w4sense(struct ccw_device *cdev, enum dev_event dev_event) /* Basic sense hasn't started. Try again. */ ccw_device_do_sense(cdev, irb); else { - CIO_MSG_EVENT(2, "Huh? 0.%x.%04x: unsolicited " + CIO_MSG_EVENT(0, "0.%x.%04x: unsolicited " "interrupt during w4sense...\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno); @@ -1169,8 +1167,10 @@ ccw_device_nop(struct ccw_device *cdev, enum dev_event dev_event) static void ccw_device_bug(struct ccw_device *cdev, enum dev_event dev_event) { - CIO_MSG_EVENT(0, "dev_jumptable[%i][%i] == NULL\n", - cdev->private->state, dev_event); + CIO_MSG_EVENT(0, "Internal state [%i][%i] not handled for device " + "0.%x.%04x\n", cdev->private->state, dev_event, + cdev->private->dev_id.ssid, + cdev->private->dev_id.devno); BUG(); } diff --git a/drivers/s390/cio/device_id.c b/drivers/s390/cio/device_id.c index dc4d87f77f6..cba7020517e 100644 --- a/drivers/s390/cio/device_id.c +++ b/drivers/s390/cio/device_id.c @@ -214,7 +214,7 @@ ccw_device_check_sense_id(struct ccw_device *cdev) * sense id information. So, for intervention required, * we use the "whack it until it talks" strategy... */ - CIO_MSG_EVENT(2, "SenseID : device %04x on Subchannel " + CIO_MSG_EVENT(0, "SenseID : device %04x on Subchannel " "0.%x.%04x reports cmd reject\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no); @@ -239,7 +239,7 @@ ccw_device_check_sense_id(struct ccw_device *cdev) lpm = to_io_private(sch)->orb.lpm; if ((lpm & sch->schib.pmcw.pim & sch->schib.pmcw.pam) != 0) - CIO_MSG_EVENT(2, "SenseID : path %02X for device %04x " + CIO_MSG_EVENT(4, "SenseID : path %02X for device %04x " "on subchannel 0.%x.%04x is " "'not operational'\n", lpm, cdev->private->dev_id.devno, diff --git a/drivers/s390/cio/device_pgid.c b/drivers/s390/cio/device_pgid.c index c52449a1f9f..ba559053402 100644 --- a/drivers/s390/cio/device_pgid.c +++ b/drivers/s390/cio/device_pgid.c @@ -79,7 +79,7 @@ __ccw_device_sense_pgid_start(struct ccw_device *cdev) /* ret is 0, -EBUSY, -EACCES or -ENODEV */ if (ret != -EACCES) return ret; - CIO_MSG_EVENT(2, "SNID - Device %04x on Subchannel " + CIO_MSG_EVENT(3, "SNID - Device %04x on Subchannel " "0.%x.%04x, lpm %02X, became 'not " "operational'\n", cdev->private->dev_id.devno, @@ -159,7 +159,7 @@ __ccw_device_check_sense_pgid(struct ccw_device *cdev) u8 lpm; lpm = to_io_private(sch)->orb.lpm; - CIO_MSG_EVENT(2, "SNID - Device %04x on Subchannel 0.%x.%04x," + CIO_MSG_EVENT(3, "SNID - Device %04x on Subchannel 0.%x.%04x," " lpm %02X, became 'not operational'\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no, lpm); @@ -275,7 +275,7 @@ __ccw_device_do_pgid(struct ccw_device *cdev, __u8 func) return ret; } /* PGID command failed on this path. */ - CIO_MSG_EVENT(2, "SPID - Device %04x on Subchannel " + CIO_MSG_EVENT(3, "SPID - Device %04x on Subchannel " "0.%x.%04x, lpm %02X, became 'not operational'\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no, cdev->private->imask); @@ -317,7 +317,7 @@ static int __ccw_device_do_nop(struct ccw_device *cdev) return ret; } /* nop command failed on this path. */ - CIO_MSG_EVENT(2, "NOP - Device %04x on Subchannel " + CIO_MSG_EVENT(3, "NOP - Device %04x on Subchannel " "0.%x.%04x, lpm %02X, became 'not operational'\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no, cdev->private->imask); @@ -362,7 +362,7 @@ __ccw_device_check_pgid(struct ccw_device *cdev) return -EAGAIN; } if (irb->scsw.cc == 3) { - CIO_MSG_EVENT(2, "SPID - Device %04x on Subchannel 0.%x.%04x," + CIO_MSG_EVENT(3, "SPID - Device %04x on Subchannel 0.%x.%04x," " lpm %02X, became 'not operational'\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no, cdev->private->imask); @@ -391,7 +391,7 @@ static int __ccw_device_check_nop(struct ccw_device *cdev) return -ETIME; } if (irb->scsw.cc == 3) { - CIO_MSG_EVENT(2, "NOP - Device %04x on Subchannel 0.%x.%04x," + CIO_MSG_EVENT(3, "NOP - Device %04x on Subchannel 0.%x.%04x," " lpm %02X, became 'not operational'\n", cdev->private->dev_id.devno, sch->schid.ssid, sch->schid.sch_no, cdev->private->imask); diff --git a/drivers/s390/s390mach.c b/drivers/s390/s390mach.c index 4d4b54277c4..5080f343ad7 100644 --- a/drivers/s390/s390mach.c +++ b/drivers/s390/s390mach.c @@ -48,10 +48,11 @@ s390_collect_crw_info(void *param) int ccode; struct semaphore *sem; unsigned int chain; + int ignore; sem = (struct semaphore *)param; repeat: - down_interruptible(sem); + ignore = down_interruptible(sem); chain = 0; while (1) { if (unlikely(chain > 1)) { diff --git a/drivers/sbus/char/bpp.c b/drivers/sbus/char/bpp.c index 4fab0c23814..b87037ec980 100644 --- a/drivers/sbus/char/bpp.c +++ b/drivers/sbus/char/bpp.c @@ -41,7 +41,7 @@ #define BPP_DELAY 100 static const unsigned BPP_MAJOR = LP_MAJOR; -static const char* dev_name = "bpp"; +static const char *bpp_dev_name = "bpp"; /* When switching from compatibility to a mode where I can read, try the following mode first. */ diff --git a/drivers/scsi/dpt_i2o.c b/drivers/scsi/dpt_i2o.c index 0fb5bf4c43a..8508816f303 100644 --- a/drivers/scsi/dpt_i2o.c +++ b/drivers/scsi/dpt_i2o.c @@ -1967,45 +1967,6 @@ cleanup: return rcode; } - -/* - * This routine returns information about the system. This does not effect - * any logic and if the info is wrong - it doesn't matter. - */ - -/* Get all the info we can not get from kernel services */ -static int adpt_system_info(void __user *buffer) -{ - sysInfo_S si; - - memset(&si, 0, sizeof(si)); - - si.osType = OS_LINUX; - si.osMajorVersion = 0; - si.osMinorVersion = 0; - si.osRevision = 0; - si.busType = SI_PCI_BUS; - si.processorFamily = DPTI_sig.dsProcessorFamily; - -#if defined __i386__ - adpt_i386_info(&si); -#elif defined (__ia64__) - adpt_ia64_info(&si); -#elif defined(__sparc__) - adpt_sparc_info(&si); -#elif defined (__alpha__) - adpt_alpha_info(&si); -#else - si.processorType = 0xff ; -#endif - if(copy_to_user(buffer, &si, sizeof(si))){ - printk(KERN_WARNING"dpti: Could not copy buffer TO user\n"); - return -EFAULT; - } - - return 0; -} - #if defined __ia64__ static void adpt_ia64_info(sysInfo_S* si) { @@ -2016,7 +1977,6 @@ static void adpt_ia64_info(sysInfo_S* si) } #endif - #if defined __sparc__ static void adpt_sparc_info(sysInfo_S* si) { @@ -2026,7 +1986,6 @@ static void adpt_sparc_info(sysInfo_S* si) si->processorType = PROC_ULTRASPARC; } #endif - #if defined __alpha__ static void adpt_alpha_info(sysInfo_S* si) { @@ -2038,7 +1997,6 @@ static void adpt_alpha_info(sysInfo_S* si) #endif #if defined __i386__ - static void adpt_i386_info(sysInfo_S* si) { // This is all the info we need for now @@ -2059,9 +2017,45 @@ static void adpt_i386_info(sysInfo_S* si) break; } } +#endif + +/* + * This routine returns information about the system. This does not effect + * any logic and if the info is wrong - it doesn't matter. + */ +/* Get all the info we can not get from kernel services */ +static int adpt_system_info(void __user *buffer) +{ + sysInfo_S si; + + memset(&si, 0, sizeof(si)); + + si.osType = OS_LINUX; + si.osMajorVersion = 0; + si.osMinorVersion = 0; + si.osRevision = 0; + si.busType = SI_PCI_BUS; + si.processorFamily = DPTI_sig.dsProcessorFamily; + +#if defined __i386__ + adpt_i386_info(&si); +#elif defined (__ia64__) + adpt_ia64_info(&si); +#elif defined(__sparc__) + adpt_sparc_info(&si); +#elif defined (__alpha__) + adpt_alpha_info(&si); +#else + si.processorType = 0xff ; #endif + if (copy_to_user(buffer, &si, sizeof(si))){ + printk(KERN_WARNING"dpti: Could not copy buffer TO user\n"); + return -EFAULT; + } + return 0; +} static int adpt_ioctl(struct inode *inode, struct file *file, uint cmd, ulong arg) diff --git a/drivers/scsi/dpti.h b/drivers/scsi/dpti.h index 924cd5a5167..337746d4604 100644 --- a/drivers/scsi/dpti.h +++ b/drivers/scsi/dpti.h @@ -316,19 +316,6 @@ static int adpt_close(struct inode *inode, struct file *file); static void adpt_delay(int millisec); #endif -#if defined __ia64__ -static void adpt_ia64_info(sysInfo_S* si); -#endif -#if defined __sparc__ -static void adpt_sparc_info(sysInfo_S* si); -#endif -#if defined __alpha__ -static void adpt_sparc_info(sysInfo_S* si); -#endif -#if defined __i386__ -static void adpt_i386_info(sysInfo_S* si); -#endif - #define PRINT_BUFFER_SIZE 512 #define HBA_FLAGS_DBG_FLAGS_MASK 0xffff0000 // Mask for debug flags diff --git a/drivers/serial/serial_core.c b/drivers/serial/serial_core.c index 1e2b9d826f6..eab03273379 100644 --- a/drivers/serial/serial_core.c +++ b/drivers/serial/serial_core.c @@ -556,7 +556,7 @@ static int uart_chars_in_buffer(struct tty_struct *tty) static void uart_flush_buffer(struct tty_struct *tty) { struct uart_state *state = tty->driver_data; - struct uart_port *port = state->port; + struct uart_port *port; unsigned long flags; /* @@ -568,6 +568,7 @@ static void uart_flush_buffer(struct tty_struct *tty) return; } + port = state->port; pr_debug("uart_flush_buffer(%d) called\n", tty->index); spin_lock_irqsave(&port->lock, flags); diff --git a/drivers/usb/host/Kconfig b/drivers/usb/host/Kconfig index 33b467a8352..1ef6df395e0 100644 --- a/drivers/usb/host/Kconfig +++ b/drivers/usb/host/Kconfig @@ -129,7 +129,7 @@ config USB_ISP1760_PCI config USB_ISP1760_OF bool "Support for the OF platform bus" - depends on USB_ISP1760_HCD && OF + depends on USB_ISP1760_HCD && PPC_OF ---help--- Enables support for the device present on the PowerPC OpenFirmware platform bus. diff --git a/drivers/usb/serial/iuu_phoenix.c b/drivers/usb/serial/iuu_phoenix.c index 8a217648b25..a01e987c7d3 100644 --- a/drivers/usb/serial/iuu_phoenix.c +++ b/drivers/usb/serial/iuu_phoenix.c @@ -643,7 +643,7 @@ static void read_buf_callback(struct urb *urb) static int iuu_bulk_write(struct usb_serial_port *port) { struct iuu_private *priv = usb_get_serial_port_data(port); - unsigned int flags; + unsigned long flags; int result; int i; char *buf_ptr = port->write_urb->transfer_buffer; @@ -694,7 +694,7 @@ static void iuu_uart_read_callback(struct urb *urb) { struct usb_serial_port *port = urb->context; struct iuu_private *priv = usb_get_serial_port_data(port); - unsigned int flags; + unsigned long flags; int status; int error = 0; int len = 0; @@ -759,7 +759,7 @@ static int iuu_uart_write(struct usb_serial_port *port, const u8 *buf, int count) { struct iuu_private *priv = usb_get_serial_port_data(port); - unsigned int flags; + unsigned long flags; dbg("%s - enter", __func__); if (count > 256) diff --git a/drivers/video/bw2.c b/drivers/video/bw2.c index 275d9dab0c6..79f85dc402d 100644 --- a/drivers/video/bw2.c +++ b/drivers/video/bw2.c @@ -329,7 +329,7 @@ static int __devinit bw2_probe(struct of_device *op, const struct of_device_id * if (!info->screen_base) goto out_unmap_regs; - bw2_blank(0, info); + bw2_blank(FB_BLANK_UNBLANK, info); bw2_init_fix(info, linebytes); diff --git a/drivers/video/cg3.c b/drivers/video/cg3.c index 010ea53978f..e31e26a6bb7 100644 --- a/drivers/video/cg3.c +++ b/drivers/video/cg3.c @@ -398,7 +398,7 @@ static int __devinit cg3_probe(struct of_device *op, if (!info->screen_base) goto out_unmap_regs; - cg3_blank(0, info); + cg3_blank(FB_BLANK_UNBLANK, info); if (!of_find_property(dp, "width", NULL)) { err = cg3_do_default_mode(par); diff --git a/drivers/video/cg6.c b/drivers/video/cg6.c index fc90db6da65..8000bccecdc 100644 --- a/drivers/video/cg6.c +++ b/drivers/video/cg6.c @@ -767,7 +767,7 @@ static int __devinit cg6_probe(struct of_device *op, cg6_bt_init(par); cg6_chip_init(info); - cg6_blank(0, info); + cg6_blank(FB_BLANK_UNBLANK, info); if (fb_alloc_cmap(&info->cmap, 256, 0)) goto out_unmap_regs; diff --git a/drivers/video/ffb.c b/drivers/video/ffb.c index 93dca3e2aa5..0f42a696d17 100644 --- a/drivers/video/ffb.c +++ b/drivers/video/ffb.c @@ -987,7 +987,7 @@ static int __devinit ffb_probe(struct of_device *op, * chosen console, it will have video outputs off in * the DAC. */ - ffb_blank(0, info); + ffb_blank(FB_BLANK_UNBLANK, info); if (fb_alloc_cmap(&info->cmap, 256, 0)) goto out_unmap_dac; diff --git a/drivers/video/leo.c b/drivers/video/leo.c index f3160fc2979..fb129928d5d 100644 --- a/drivers/video/leo.c +++ b/drivers/video/leo.c @@ -601,7 +601,7 @@ static int __devinit leo_probe(struct of_device *op, const struct of_device_id * leo_init_wids(info); leo_init_hw(info); - leo_blank(0, info); + leo_blank(FB_BLANK_UNBLANK, info); if (fb_alloc_cmap(&info->cmap, 256, 0)) goto out_unmap_regs; diff --git a/drivers/video/p9100.c b/drivers/video/p9100.c index c95874fe907..676ffb06d1c 100644 --- a/drivers/video/p9100.c +++ b/drivers/video/p9100.c @@ -295,7 +295,7 @@ static int __devinit p9100_probe(struct of_device *op, const struct of_device_id if (!info->screen_base) goto out_unmap_regs; - p9100_blank(0, info); + p9100_blank(FB_BLANK_UNBLANK, info); if (fb_alloc_cmap(&info->cmap, 256, 0)) goto out_unmap_screen; diff --git a/drivers/video/tcx.c b/drivers/video/tcx.c index a7177430577..44e8c27ed0f 100644 --- a/drivers/video/tcx.c +++ b/drivers/video/tcx.c @@ -84,7 +84,7 @@ struct tcx_tec { struct tcx_thc { u32 thc_rev; - u32 thc_pad0[511]; + u32 thc_pad0[511]; u32 thc_hs; /* hsync timing */ u32 thc_hsdvs; u32 thc_hd; @@ -126,10 +126,10 @@ struct tcx_par { }; /* Reset control plane so that WID is 8-bit plane. */ -static void __tcx_set_control_plane (struct tcx_par *par) +static void __tcx_set_control_plane(struct tcx_par *par) { u32 __iomem *p, *pend; - + if (par->lowdepth) return; @@ -143,8 +143,8 @@ static void __tcx_set_control_plane (struct tcx_par *par) sbus_writel(tmp, p); } } - -static void tcx_reset (struct fb_info *info) + +static void tcx_reset(struct fb_info *info) { struct tcx_par *par = (struct tcx_par *) info->par; unsigned long flags; @@ -365,7 +365,8 @@ static void tcx_unmap_regs(struct of_device *op, struct fb_info *info, info->screen_base, par->fbsize); } -static int __devinit tcx_init_one(struct of_device *op) +static int __devinit tcx_probe(struct of_device *op, + const struct of_device_id *match) { struct device_node *dp = op->node; struct fb_info *info; @@ -488,13 +489,6 @@ out_err: return err; } -static int __devinit tcx_probe(struct of_device *dev, const struct of_device_id *match) -{ - struct of_device *op = to_of_device(&dev->dev); - - return tcx_init_one(op); -} - static int __devexit tcx_remove(struct of_device *op) { struct fb_info *info = dev_get_drvdata(&op->dev); diff --git a/fs/affs/affs.h b/fs/affs/affs.h index d5bd497ab9c..223b1917093 100644 --- a/fs/affs/affs.h +++ b/fs/affs/affs.h @@ -48,7 +48,7 @@ struct affs_ext_key { * affs fs inode data in memory */ struct affs_inode_info { - u32 i_opencnt; + atomic_t i_opencnt; struct semaphore i_link_lock; /* Protects internal inode access. */ struct semaphore i_ext_lock; /* Protects internal inode access. */ #define i_hash_lock i_ext_lock @@ -170,8 +170,6 @@ extern int affs_rename(struct inode *old_dir, struct dentry *old_dentry, extern unsigned long affs_parent_ino(struct inode *dir); extern struct inode *affs_new_inode(struct inode *dir); extern int affs_notify_change(struct dentry *dentry, struct iattr *attr); -extern void affs_put_inode(struct inode *inode); -extern void affs_drop_inode(struct inode *inode); extern void affs_delete_inode(struct inode *inode); extern void affs_clear_inode(struct inode *inode); extern struct inode *affs_iget(struct super_block *sb, diff --git a/fs/affs/file.c b/fs/affs/file.c index 1a4f092f24e..6eac7bdeec9 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -48,8 +48,9 @@ affs_file_open(struct inode *inode, struct file *filp) { if (atomic_read(&filp->f_count) != 1) return 0; - pr_debug("AFFS: open(%d)\n", AFFS_I(inode)->i_opencnt); - AFFS_I(inode)->i_opencnt++; + pr_debug("AFFS: open(%lu,%d)\n", + inode->i_ino, atomic_read(&AFFS_I(inode)->i_opencnt)); + atomic_inc(&AFFS_I(inode)->i_opencnt); return 0; } @@ -58,10 +59,16 @@ affs_file_release(struct inode *inode, struct file *filp) { if (atomic_read(&filp->f_count) != 0) return 0; - pr_debug("AFFS: release(%d)\n", AFFS_I(inode)->i_opencnt); - AFFS_I(inode)->i_opencnt--; - if (!AFFS_I(inode)->i_opencnt) + pr_debug("AFFS: release(%lu, %d)\n", + inode->i_ino, atomic_read(&AFFS_I(inode)->i_opencnt)); + + if (atomic_dec_and_test(&AFFS_I(inode)->i_opencnt)) { + mutex_lock(&inode->i_mutex); + if (inode->i_size != AFFS_I(inode)->mmu_private) + affs_truncate(inode); affs_free_prealloc(inode); + mutex_unlock(&inode->i_mutex); + } return 0; } @@ -180,7 +187,7 @@ affs_get_extblock(struct inode *inode, u32 ext) /* inline the simplest case: same extended block as last time */ struct buffer_head *bh = AFFS_I(inode)->i_ext_bh; if (ext == AFFS_I(inode)->i_ext_last) - atomic_inc(&bh->b_count); + get_bh(bh); else /* we have to do more (not inlined) */ bh = affs_get_extblock_slow(inode, ext); @@ -306,7 +313,7 @@ store_ext: affs_brelse(AFFS_I(inode)->i_ext_bh); AFFS_I(inode)->i_ext_last = ext; AFFS_I(inode)->i_ext_bh = bh; - atomic_inc(&bh->b_count); + get_bh(bh); return bh; @@ -324,7 +331,6 @@ affs_get_block(struct inode *inode, sector_t block, struct buffer_head *bh_resul pr_debug("AFFS: get_block(%u, %lu)\n", (u32)inode->i_ino, (unsigned long)block); - BUG_ON(block > (sector_t)0x7fffffffUL); if (block >= AFFS_I(inode)->i_blkcnt) { @@ -827,6 +833,8 @@ affs_truncate(struct inode *inode) res = mapping->a_ops->write_begin(NULL, mapping, size, 0, 0, &page, &fsdata); if (!res) res = mapping->a_ops->write_end(NULL, mapping, size, 0, 0, page, fsdata); + else + inode->i_size = AFFS_I(inode)->mmu_private; mark_inode_dirty(inode); return; } else if (inode->i_size == AFFS_I(inode)->mmu_private) @@ -862,6 +870,7 @@ affs_truncate(struct inode *inode) blk++; } else AFFS_HEAD(ext_bh)->first_data = 0; + AFFS_HEAD(ext_bh)->block_count = cpu_to_be32(i); size = AFFS_SB(sb)->s_hashsize; if (size > blkcnt - blk + i) size = blkcnt - blk + i; diff --git a/fs/affs/inode.c b/fs/affs/inode.c index 27fe6cbe43a..a13b334a391 100644 --- a/fs/affs/inode.c +++ b/fs/affs/inode.c @@ -58,7 +58,7 @@ struct inode *affs_iget(struct super_block *sb, unsigned long ino) AFFS_I(inode)->i_extcnt = 1; AFFS_I(inode)->i_ext_last = ~1; AFFS_I(inode)->i_protect = prot; - AFFS_I(inode)->i_opencnt = 0; + atomic_set(&AFFS_I(inode)->i_opencnt, 0); AFFS_I(inode)->i_blkcnt = 0; AFFS_I(inode)->i_lc = NULL; AFFS_I(inode)->i_lc_size = 0; @@ -108,8 +108,6 @@ struct inode *affs_iget(struct super_block *sb, unsigned long ino) inode->i_mode |= S_IFDIR; } else inode->i_mode = S_IRUGO | S_IXUGO | S_IWUSR | S_IFDIR; - if (tail->link_chain) - inode->i_nlink = 2; /* Maybe it should be controlled by mount parameter? */ //inode->i_mode |= S_ISVTX; inode->i_op = &affs_dir_inode_operations; @@ -245,31 +243,12 @@ out: } void -affs_put_inode(struct inode *inode) -{ - pr_debug("AFFS: put_inode(ino=%lu, nlink=%u)\n", inode->i_ino, inode->i_nlink); - affs_free_prealloc(inode); -} - -void -affs_drop_inode(struct inode *inode) -{ - mutex_lock(&inode->i_mutex); - if (inode->i_size != AFFS_I(inode)->mmu_private) - affs_truncate(inode); - mutex_unlock(&inode->i_mutex); - - generic_drop_inode(inode); -} - -void affs_delete_inode(struct inode *inode) { pr_debug("AFFS: delete_inode(ino=%lu, nlink=%u)\n", inode->i_ino, inode->i_nlink); truncate_inode_pages(&inode->i_data, 0); inode->i_size = 0; - if (S_ISREG(inode->i_mode)) - affs_truncate(inode); + affs_truncate(inode); clear_inode(inode); affs_free_block(inode->i_sb, inode->i_ino); } @@ -277,9 +256,12 @@ affs_delete_inode(struct inode *inode) void affs_clear_inode(struct inode *inode) { - unsigned long cache_page = (unsigned long) AFFS_I(inode)->i_lc; + unsigned long cache_page; pr_debug("AFFS: clear_inode(ino=%lu, nlink=%u)\n", inode->i_ino, inode->i_nlink); + + affs_free_prealloc(inode); + cache_page = (unsigned long)AFFS_I(inode)->i_lc; if (cache_page) { pr_debug("AFFS: freeing ext cache\n"); AFFS_I(inode)->i_lc = NULL; @@ -316,7 +298,7 @@ affs_new_inode(struct inode *dir) inode->i_ino = block; inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - AFFS_I(inode)->i_opencnt = 0; + atomic_set(&AFFS_I(inode)->i_opencnt, 0); AFFS_I(inode)->i_blkcnt = 0; AFFS_I(inode)->i_lc = NULL; AFFS_I(inode)->i_lc_size = 0; @@ -369,12 +351,12 @@ affs_add_entry(struct inode *dir, struct inode *inode, struct dentry *dentry, s3 switch (type) { case ST_LINKFILE: case ST_LINKDIR: - inode_bh = bh; retval = -ENOSPC; block = affs_alloc_block(dir, dir->i_ino); if (!block) goto err; retval = -EIO; + inode_bh = bh; bh = affs_getzeroblk(sb, block); if (!bh) goto err; diff --git a/fs/affs/namei.c b/fs/affs/namei.c index 2218f1ee71c..cfcf1b6cf82 100644 --- a/fs/affs/namei.c +++ b/fs/affs/namei.c @@ -234,7 +234,8 @@ affs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) int affs_unlink(struct inode *dir, struct dentry *dentry) { - pr_debug("AFFS: unlink(dir=%d, \"%.*s\")\n", (u32)dir->i_ino, + pr_debug("AFFS: unlink(dir=%d, %lu \"%.*s\")\n", (u32)dir->i_ino, + dentry->d_inode->i_ino, (int)dentry->d_name.len, dentry->d_name.name); return affs_remove_header(dentry); @@ -302,7 +303,8 @@ affs_mkdir(struct inode *dir, struct dentry *dentry, int mode) int affs_rmdir(struct inode *dir, struct dentry *dentry) { - pr_debug("AFFS: rmdir(dir=%u, \"%.*s\")\n", (u32)dir->i_ino, + pr_debug("AFFS: rmdir(dir=%u, %lu \"%.*s\")\n", (u32)dir->i_ino, + dentry->d_inode->i_ino, (int)dentry->d_name.len, dentry->d_name.name); return affs_remove_header(dentry); diff --git a/fs/affs/super.c b/fs/affs/super.c index 01d25d53254..d214837d5e4 100644 --- a/fs/affs/super.c +++ b/fs/affs/super.c @@ -71,12 +71,18 @@ static struct kmem_cache * affs_inode_cachep; static struct inode *affs_alloc_inode(struct super_block *sb) { - struct affs_inode_info *ei; - ei = (struct affs_inode_info *)kmem_cache_alloc(affs_inode_cachep, GFP_KERNEL); - if (!ei) + struct affs_inode_info *i; + + i = kmem_cache_alloc(affs_inode_cachep, GFP_KERNEL); + if (!i) return NULL; - ei->vfs_inode.i_version = 1; - return &ei->vfs_inode; + + i->vfs_inode.i_version = 1; + i->i_lc = NULL; + i->i_ext_bh = NULL; + i->i_pa_cnt = 0; + + return &i->vfs_inode; } static void affs_destroy_inode(struct inode *inode) @@ -114,8 +120,6 @@ static const struct super_operations affs_sops = { .alloc_inode = affs_alloc_inode, .destroy_inode = affs_destroy_inode, .write_inode = affs_write_inode, - .put_inode = affs_put_inode, - .drop_inode = affs_drop_inode, .delete_inode = affs_delete_inode, .clear_inode = affs_clear_inode, .put_super = affs_put_super, @@ -158,7 +158,7 @@ struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) bio_init(bio); if (likely(nr_iovecs)) { - unsigned long idx = 0; /* shut up gcc */ + unsigned long uninitialized_var(idx); bvl = bvec_alloc_bs(gfp_mask, nr_iovecs, &idx, bs); if (unlikely(!bvl)) { @@ -963,6 +963,7 @@ static void bio_copy_kern_endio(struct bio *bio, int err) * @data: pointer to buffer to copy * @len: length in bytes * @gfp_mask: allocation flags for bio and page allocation + * @reading: data direction is READ * * copy the kernel address into a bio suitable for io to a block * device. Returns an error pointer in case of error. diff --git a/fs/inode.c b/fs/inode.c index bf647813042..c36d9480335 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1149,13 +1149,8 @@ static inline void iput_final(struct inode *inode) void iput(struct inode *inode) { if (inode) { - const struct super_operations *op = inode->i_sb->s_op; - BUG_ON(inode->i_state == I_CLEAR); - if (op && op->put_inode) - op->put_inode(inode); - if (atomic_dec_and_lock(&inode->i_count, &inode_lock)) iput_final(inode); } diff --git a/fs/locks.c b/fs/locks.c index 663c069b59b..0ac6b92cb0b 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -1753,6 +1753,7 @@ int fcntl_setlk(unsigned int fd, struct file *filp, unsigned int cmd, struct file_lock *file_lock = locks_alloc_lock(); struct flock flock; struct inode *inode; + struct file *f; int error; if (file_lock == NULL) @@ -1825,7 +1826,15 @@ again: * Attempt to detect a close/fcntl race and recover by * releasing the lock that was just acquired. */ - if (!error && fcheck(fd) != filp && flock.l_type != F_UNLCK) { + /* + * we need that spin_lock here - it prevents reordering between + * update of inode->i_flock and check for it done in close(). + * rcu_read_lock() wouldn't do. + */ + spin_lock(¤t->files->file_lock); + f = fcheck(fd); + spin_unlock(¤t->files->file_lock); + if (!error && f != filp && flock.l_type != F_UNLCK) { flock.l_type = F_UNLCK; goto again; } @@ -1881,6 +1890,7 @@ int fcntl_setlk64(unsigned int fd, struct file *filp, unsigned int cmd, struct file_lock *file_lock = locks_alloc_lock(); struct flock64 flock; struct inode *inode; + struct file *f; int error; if (file_lock == NULL) @@ -1953,7 +1963,10 @@ again: * Attempt to detect a close/fcntl race and recover by * releasing the lock that was just acquired. */ - if (!error && fcheck(fd) != filp && flock.l_type != F_UNLCK) { + spin_lock(¤t->files->file_lock); + f = fcheck(fd); + spin_unlock(¤t->files->file_lock); + if (!error && f != filp && flock.l_type != F_UNLCK) { flock.l_type = F_UNLCK; goto again; } diff --git a/fs/pipe.c b/fs/pipe.c index 3499f9ff631..ec228bc9f88 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -17,6 +17,7 @@ #include <linux/highmem.h> #include <linux/pagemap.h> #include <linux/audit.h> +#include <linux/syscalls.h> #include <asm/uaccess.h> #include <asm/ioctls.h> @@ -1086,8 +1087,11 @@ asmlinkage long __weak sys_pipe(int __user *fildes) error = do_pipe(fd); if (!error) { - if (copy_to_user(fildes, fd, sizeof(fd))) + if (copy_to_user(fildes, fd, sizeof(fd))) { + sys_close(fd[0]); + sys_close(fd[1]); error = -EFAULT; + } } return error; } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index e2b8e769f51..88717c0f941 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -5,11 +5,9 @@ #include <linux/highmem.h> #include <linux/ptrace.h> #include <linux/pagemap.h> -#include <linux/ptrace.h> #include <linux/mempolicy.h> #include <linux/swap.h> #include <linux/swapops.h> -#include <linux/seq_file.h> #include <asm/elf.h> #include <asm/uaccess.h> diff --git a/fs/splice.c b/fs/splice.c index 633f58ebfb7..78150038b58 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -811,24 +811,19 @@ generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out, { struct address_space *mapping = out->f_mapping; struct inode *inode = mapping->host; - int killsuid, killpriv; + struct splice_desc sd = { + .total_len = len, + .flags = flags, + .pos = *ppos, + .u.file = out, + }; ssize_t ret; - int err = 0; - - killpriv = security_inode_need_killpriv(out->f_path.dentry); - killsuid = should_remove_suid(out->f_path.dentry); - if (unlikely(killsuid || killpriv)) { - mutex_lock(&inode->i_mutex); - if (killpriv) - err = security_inode_killpriv(out->f_path.dentry); - if (!err && killsuid) - err = __remove_suid(out->f_path.dentry, killsuid); - mutex_unlock(&inode->i_mutex); - if (err) - return err; - } - ret = splice_from_pipe(pipe, out, ppos, len, flags, pipe_to_file); + inode_double_lock(inode, pipe->inode); + ret = remove_suid(out->f_path.dentry); + if (likely(!ret)) + ret = __splice_from_pipe(pipe, &sd, pipe_to_file); + inode_double_unlock(inode, pipe->inode); if (ret > 0) { unsigned long nr_pages; @@ -840,6 +835,8 @@ generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out, * sync it. */ if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) { + int err; + mutex_lock(&inode->i_mutex); err = generic_osync_inode(inode, mapping, OSYNC_METADATA|OSYNC_DATA); @@ -1075,7 +1072,7 @@ long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, ret = splice_direct_to_actor(in, &sd, direct_splice_actor); if (ret > 0) - *ppos = sd.pos; + *ppos += ret; return ret; } diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 2b34c8ca6c8..d3231947db1 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -32,6 +32,7 @@ #include <linux/buffer_head.h> #include <linux/sched.h> #include <linux/crc-itu-t.h> +#include <linux/exportfs.h> static inline int udf_match(int len1, const char *name1, int len2, const char *name2) @@ -158,6 +159,8 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, sector_t offset; struct extent_position epos = {}; struct udf_inode_info *dinfo = UDF_I(dir); + int isdotdot = dentry->d_name.len == 2 && + dentry->d_name.name[0] == '.' && dentry->d_name.name[1] == '.'; size = udf_ext0_offset(dir) + dir->i_size; f_pos = udf_ext0_offset(dir); @@ -225,6 +228,12 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, continue; } + if ((cfi->fileCharacteristics & FID_FILE_CHAR_PARENT) && + isdotdot) { + brelse(epos.bh); + return fi; + } + if (!lfi) continue; @@ -286,9 +295,8 @@ static struct dentry *udf_lookup(struct inode *dir, struct dentry *dentry, } } unlock_kernel(); - d_add(dentry, inode); - return NULL; + return d_splice_alias(inode, dentry); } static struct fileIdentDesc *udf_add_entry(struct inode *dir, @@ -307,7 +315,7 @@ static struct fileIdentDesc *udf_add_entry(struct inode *dir, uint16_t liu; int block; kernel_lb_addr eloc; - uint32_t elen; + uint32_t elen = 0; sector_t offset; struct extent_position epos = {}; struct udf_inode_info *dinfo; @@ -398,7 +406,8 @@ static struct fileIdentDesc *udf_add_entry(struct inode *dir, } add: - if (dinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB) { + /* Is there any extent whose size we need to round up? */ + if (dinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB && elen) { elen = (elen + sb->s_blocksize - 1) & ~(sb->s_blocksize - 1); if (dinfo->i_alloc_type == ICBTAG_FLAG_AD_SHORT) epos.offset -= sizeof(short_ad); @@ -1232,6 +1241,134 @@ end_rename: return retval; } +static struct dentry *udf_get_parent(struct dentry *child) +{ + struct dentry *parent; + struct inode *inode = NULL; + struct dentry dotdot; + struct fileIdentDesc cfi; + struct udf_fileident_bh fibh; + + dotdot.d_name.name = ".."; + dotdot.d_name.len = 2; + + lock_kernel(); + if (!udf_find_entry(child->d_inode, &dotdot, &fibh, &cfi)) + goto out_unlock; + + if (fibh.sbh != fibh.ebh) + brelse(fibh.ebh); + brelse(fibh.sbh); + + inode = udf_iget(child->d_inode->i_sb, + lelb_to_cpu(cfi.icb.extLocation)); + if (!inode) + goto out_unlock; + unlock_kernel(); + + parent = d_alloc_anon(inode); + if (!parent) { + iput(inode); + parent = ERR_PTR(-ENOMEM); + } + + return parent; +out_unlock: + unlock_kernel(); + return ERR_PTR(-EACCES); +} + + +static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block, + u16 partref, __u32 generation) +{ + struct inode *inode; + struct dentry *result; + kernel_lb_addr loc; + + if (block == 0) + return ERR_PTR(-ESTALE); + + loc.logicalBlockNum = block; + loc.partitionReferenceNum = partref; + inode = udf_iget(sb, loc); + + if (inode == NULL) + return ERR_PTR(-ENOMEM); + + if (generation && inode->i_generation != generation) { + iput(inode); + return ERR_PTR(-ESTALE); + } + result = d_alloc_anon(inode); + if (!result) { + iput(inode); + return ERR_PTR(-ENOMEM); + } + return result; +} + +static struct dentry *udf_fh_to_dentry(struct super_block *sb, + struct fid *fid, int fh_len, int fh_type) +{ + if ((fh_len != 3 && fh_len != 5) || + (fh_type != FILEID_UDF_WITH_PARENT && + fh_type != FILEID_UDF_WITHOUT_PARENT)) + return NULL; + + return udf_nfs_get_inode(sb, fid->udf.block, fid->udf.partref, + fid->udf.generation); +} + +static struct dentry *udf_fh_to_parent(struct super_block *sb, + struct fid *fid, int fh_len, int fh_type) +{ + if (fh_len != 5 || fh_type != FILEID_UDF_WITH_PARENT) + return NULL; + + return udf_nfs_get_inode(sb, fid->udf.parent_block, + fid->udf.parent_partref, + fid->udf.parent_generation); +} +static int udf_encode_fh(struct dentry *de, __u32 *fh, int *lenp, + int connectable) +{ + int len = *lenp; + struct inode *inode = de->d_inode; + kernel_lb_addr location = UDF_I(inode)->i_location; + struct fid *fid = (struct fid *)fh; + int type = FILEID_UDF_WITHOUT_PARENT; + + if (len < 3 || (connectable && len < 5)) + return 255; + + *lenp = 3; + fid->udf.block = location.logicalBlockNum; + fid->udf.partref = location.partitionReferenceNum; + fid->udf.generation = inode->i_generation; + + if (connectable && !S_ISDIR(inode->i_mode)) { + spin_lock(&de->d_lock); + inode = de->d_parent->d_inode; + location = UDF_I(inode)->i_location; + fid->udf.parent_block = location.logicalBlockNum; + fid->udf.parent_partref = location.partitionReferenceNum; + fid->udf.parent_generation = inode->i_generation; + spin_unlock(&de->d_lock); + *lenp = 5; + type = FILEID_UDF_WITH_PARENT; + } + + return type; +} + +const struct export_operations udf_export_ops = { + .encode_fh = udf_encode_fh, + .fh_to_dentry = udf_fh_to_dentry, + .fh_to_parent = udf_fh_to_parent, + .get_parent = udf_get_parent, +}; + const struct inode_operations udf_dir_inode_operations = { .lookup = udf_lookup, .create = udf_create, diff --git a/fs/udf/partition.c b/fs/udf/partition.c index 63610f026ae..96dfd207c3d 100644 --- a/fs/udf/partition.c +++ b/fs/udf/partition.c @@ -27,8 +27,8 @@ #include <linux/slab.h> #include <linux/buffer_head.h> -inline uint32_t udf_get_pblock(struct super_block *sb, uint32_t block, - uint16_t partition, uint32_t offset) +uint32_t udf_get_pblock(struct super_block *sb, uint32_t block, + uint16_t partition, uint32_t offset) { struct udf_sb_info *sbi = UDF_SB(sb); struct udf_part_map *map; diff --git a/fs/udf/super.c b/fs/udf/super.c index 9fb18a340fc..7a5f69be6ac 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1933,6 +1933,7 @@ static int udf_fill_super(struct super_block *sb, void *options, int silent) /* Fill in the rest of the superblock */ sb->s_op = &udf_sb_ops; + sb->s_export_op = &udf_export_ops; sb->dq_op = NULL; sb->s_dirt = 0; sb->s_magic = UDF_SUPER_MAGIC; diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h index f3f45d02927..8fa9c2d7091 100644 --- a/fs/udf/udfdecl.h +++ b/fs/udf/udfdecl.h @@ -73,6 +73,7 @@ struct task_struct; struct buffer_head; struct super_block; +extern const struct export_operations udf_export_ops; extern const struct inode_operations udf_dir_inode_operations; extern const struct file_operations udf_dir_operations; extern const struct inode_operations udf_file_inode_operations; diff --git a/include/asm-mn10300/processor.h b/include/asm-mn10300/processor.h index f1b081f5346..73239271873 100644 --- a/include/asm-mn10300/processor.h +++ b/include/asm-mn10300/processor.h @@ -58,7 +58,7 @@ extern struct mn10300_cpuinfo boot_cpu_data; extern void identify_cpu(struct mn10300_cpuinfo *); extern void print_cpu_info(struct mn10300_cpuinfo *); extern void dodgy_tsc(void); -#define cpu_relax() do {} while (0) +#define cpu_relax() barrier() /* * User space process size: 1.75GB (default). diff --git a/include/asm-s390/kvm_host.h b/include/asm-s390/kvm_host.h index f8204a4f2e0..18cbd8a3979 100644 --- a/include/asm-s390/kvm_host.h +++ b/include/asm-s390/kvm_host.h @@ -104,6 +104,7 @@ struct sie_block { struct kvm_vcpu_stat { u32 exit_userspace; + u32 exit_null; u32 exit_external_request; u32 exit_external_interrupt; u32 exit_stop_request; diff --git a/include/asm-s390/page.h b/include/asm-s390/page.h index f0f4579eac1..12fd9c4f0f1 100644 --- a/include/asm-s390/page.h +++ b/include/asm-s390/page.h @@ -125,6 +125,17 @@ page_get_storage_key(unsigned long addr) return skey; } +#ifdef CONFIG_PAGE_STATES + +struct page; +void arch_free_page(struct page *page, int order); +void arch_alloc_page(struct page *page, int order); + +#define HAVE_ARCH_FREE_PAGE +#define HAVE_ARCH_ALLOC_PAGE + +#endif + #endif /* !__ASSEMBLY__ */ /* to align the pointer to the (next) page boundary */ diff --git a/include/asm-s390/ptrace.h b/include/asm-s390/ptrace.h index 441d7c26085..d7d4e2eb3e6 100644 --- a/include/asm-s390/ptrace.h +++ b/include/asm-s390/ptrace.h @@ -471,6 +471,8 @@ struct task_struct; extern void user_enable_single_step(struct task_struct *); extern void user_disable_single_step(struct task_struct *); +#define __ARCH_WANT_COMPAT_SYS_PTRACE + #define user_mode(regs) (((regs)->psw.mask & PSW_MASK_PSTATE) != 0) #define instruction_pointer(regs) ((regs)->psw.addr & PSW_ADDR_INSN) #define regs_return_value(regs)((regs)->gprs[2]) diff --git a/include/asm-s390/system.h b/include/asm-s390/system.h index c819ae25a84..e0d4500d5f9 100644 --- a/include/asm-s390/system.h +++ b/include/asm-s390/system.h @@ -116,6 +116,12 @@ extern void pfault_fini(void); #define pfault_fini() do { } while (0) #endif /* CONFIG_PFAULT */ +#ifdef CONFIG_PAGE_STATES +extern void cmma_init(void); +#else +static inline void cmma_init(void) { } +#endif + #define finish_arch_switch(prev) do { \ set_fs(current->thread.mm_segment); \ account_vtime(prev); \ diff --git a/include/asm-x86/pgtable_32.h b/include/asm-x86/pgtable_32.h index 577ab79c4c2..d7f0403bbec 100644 --- a/include/asm-x86/pgtable_32.h +++ b/include/asm-x86/pgtable_32.h @@ -88,14 +88,7 @@ extern unsigned long pg0[]; /* To avoid harmful races, pmd_none(x) should check only the lower when PAE */ #define pmd_none(x) (!(unsigned long)pmd_val((x))) #define pmd_present(x) (pmd_val((x)) & _PAGE_PRESENT) - -extern int pmd_bad(pmd_t pmd); - -#define pmd_bad_v1(x) \ - (_KERNPG_TABLE != (pmd_val((x)) & ~(PAGE_MASK | _PAGE_USER))) -#define pmd_bad_v2(x) \ - (_KERNPG_TABLE != (pmd_val((x)) & ~(PAGE_MASK | _PAGE_USER | \ - _PAGE_PSE | _PAGE_NX))) +#define pmd_bad(x) ((pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE) #define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT)) diff --git a/include/asm-x86/pgtable_64.h b/include/asm-x86/pgtable_64.h index a3bbf8766c1..efe83dcbd41 100644 --- a/include/asm-x86/pgtable_64.h +++ b/include/asm-x86/pgtable_64.h @@ -158,14 +158,12 @@ static inline unsigned long pgd_bad(pgd_t pgd) static inline unsigned long pud_bad(pud_t pud) { - return pud_val(pud) & - ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX); + return pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER); } static inline unsigned long pmd_bad(pmd_t pmd) { - return pmd_val(pmd) & - ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX); + return pmd_val(pmd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER); } #define pte_none(x) (!pte_val((x))) diff --git a/include/linux/exportfs.h b/include/linux/exportfs.h index de8387b7ceb..f5abd130663 100644 --- a/include/linux/exportfs.h +++ b/include/linux/exportfs.h @@ -33,6 +33,19 @@ enum fid_type { * 32 bit parent directory inode number. */ FILEID_INO32_GEN_PARENT = 2, + + /* + * 32 bit block number, 16 bit partition reference, + * 16 bit unused, 32 bit generation number. + */ + FILEID_UDF_WITHOUT_PARENT = 0x51, + + /* + * 32 bit block number, 16 bit partition reference, + * 16 bit unused, 32 bit generation number, + * 32 bit parent block number, 32 bit parent generation number + */ + FILEID_UDF_WITH_PARENT = 0x52, }; struct fid { @@ -43,6 +56,14 @@ struct fid { u32 parent_ino; u32 parent_gen; } i32; + struct { + u32 block; + u16 partref; + u16 parent_partref; + u32 generation; + u32 parent_block; + u32 parent_generation; + } udf; __u32 raw[0]; }; }; diff --git a/include/linux/fs.h b/include/linux/fs.h index a1ba005d08e..f413085f748 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1289,17 +1289,12 @@ extern ssize_t vfs_readv(struct file *, const struct iovec __user *, extern ssize_t vfs_writev(struct file *, const struct iovec __user *, unsigned long, loff_t *); -/* - * NOTE: write_inode, delete_inode, clear_inode, put_inode can be called - * without the big kernel lock held in all filesystems. - */ struct super_operations { struct inode *(*alloc_inode)(struct super_block *sb); void (*destroy_inode)(struct inode *); void (*dirty_inode) (struct inode *); int (*write_inode) (struct inode *, int); - void (*put_inode) (struct inode *); void (*drop_inode) (struct inode *); void (*delete_inode) (struct inode *); void (*put_super) (struct super_block *); @@ -1821,7 +1816,6 @@ extern void iget_failed(struct inode *); extern void clear_inode(struct inode *); extern void destroy_inode(struct inode *); extern struct inode *new_inode(struct super_block *); -extern int __remove_suid(struct dentry *, int); extern int should_remove_suid(struct dentry *); extern int remove_suid(struct dentry *); diff --git a/include/linux/genhd.h b/include/linux/genhd.h index ecd2bf63fc8..e9874e7fcdf 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -178,17 +178,17 @@ static inline struct hd_struct *get_part(struct gendisk *gendiskp, static inline void disk_stat_set_all(struct gendisk *gendiskp, int value) { int i; + for_each_possible_cpu(i) memset(per_cpu_ptr(gendiskp->dkstats, i), value, - sizeof (struct disk_stats)); + sizeof(struct disk_stats)); } #define __part_stat_add(part, field, addnd) \ (per_cpu_ptr(part->dkstats, smp_processor_id())->field += addnd) -#define __all_stat_add(gendiskp, field, addnd, sector) \ +#define __all_stat_add(gendiskp, part, field, addnd, sector) \ ({ \ - struct hd_struct *part = get_part(gendiskp, sector); \ if (part) \ __part_stat_add(part, field, addnd); \ __disk_stat_add(gendiskp, field, addnd); \ @@ -203,11 +203,13 @@ static inline void disk_stat_set_all(struct gendisk *gendiskp, int value) { res; \ }) -static inline void part_stat_set_all(struct hd_struct *part, int value) { +static inline void part_stat_set_all(struct hd_struct *part, int value) +{ int i; + for_each_possible_cpu(i) memset(per_cpu_ptr(part->dkstats, i), value, - sizeof(struct disk_stats)); + sizeof(struct disk_stats)); } #else /* !CONFIG_SMP */ @@ -223,9 +225,8 @@ static inline void disk_stat_set_all(struct gendisk *gendiskp, int value) #define __part_stat_add(part, field, addnd) \ (part->dkstats.field += addnd) -#define __all_stat_add(gendiskp, field, addnd, sector) \ +#define __all_stat_add(gendiskp, part, field, addnd, sector) \ ({ \ - struct hd_struct *part = get_part(gendiskp, sector); \ if (part) \ part->dkstats.field += addnd; \ __disk_stat_add(gendiskp, field, addnd); \ @@ -276,10 +277,10 @@ static inline void part_stat_set_all(struct hd_struct *part, int value) #define part_stat_sub(gendiskp, field, subnd) \ part_stat_add(gendiskp, field, -subnd) -#define all_stat_add(gendiskp, field, addnd, sector) \ +#define all_stat_add(gendiskp, part, field, addnd, sector) \ do { \ preempt_disable(); \ - __all_stat_add(gendiskp, field, addnd, sector); \ + __all_stat_add(gendiskp, part, field, addnd, sector); \ preempt_enable(); \ } while (0) @@ -288,15 +289,15 @@ static inline void part_stat_set_all(struct hd_struct *part, int value) #define all_stat_dec(gendiskp, field, sector) \ all_stat_add(gendiskp, field, -1, sector) -#define __all_stat_inc(gendiskp, field, sector) \ - __all_stat_add(gendiskp, field, 1, sector) -#define all_stat_inc(gendiskp, field, sector) \ - all_stat_add(gendiskp, field, 1, sector) +#define __all_stat_inc(gendiskp, part, field, sector) \ + __all_stat_add(gendiskp, part, field, 1, sector) +#define all_stat_inc(gendiskp, part, field, sector) \ + all_stat_add(gendiskp, part, field, 1, sector) -#define __all_stat_sub(gendiskp, field, subnd, sector) \ - __all_stat_add(gendiskp, field, -subnd, sector) -#define all_stat_sub(gendiskp, field, subnd, sector) \ - all_stat_add(gendiskp, field, -subnd, sector) +#define __all_stat_sub(gendiskp, part, field, subnd, sector) \ + __all_stat_add(gendiskp, part, field, -subnd, sector) +#define all_stat_sub(gendiskp, part, field, subnd, sector) \ + all_stat_add(gendiskp, part, field, -subnd, sector) /* Inlines to alloc and free disk stats in struct gendisk */ #ifdef CONFIG_SMP diff --git a/include/linux/ioprio.h b/include/linux/ioprio.h index 2a3bb1bb743..f98a656b17e 100644 --- a/include/linux/ioprio.h +++ b/include/linux/ioprio.h @@ -68,6 +68,20 @@ static inline int task_nice_ioprio(struct task_struct *task) } /* + * This is for the case where the task hasn't asked for a specific IO class. + * Check for idle and rt task process, and return appropriate IO class. + */ +static inline int task_nice_ioclass(struct task_struct *task) +{ + if (task->policy == SCHED_IDLE) + return IOPRIO_CLASS_IDLE; + else if (task->policy == SCHED_FIFO || task->policy == SCHED_RR) + return IOPRIO_CLASS_RT; + else + return IOPRIO_CLASS_BE; +} + +/* * For inheritance, return the highest of the two given priorities */ extern int ioprio_best(unsigned short aprio, unsigned short bprio); diff --git a/include/linux/libata.h b/include/linux/libata.h index d1dfe872ee3..7e206da1fbf 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -1039,6 +1039,7 @@ extern void ata_eh_thaw_port(struct ata_port *ap); extern void ata_eh_qc_complete(struct ata_queued_cmd *qc); extern void ata_eh_qc_retry(struct ata_queued_cmd *qc); +extern void ata_eh_analyze_ncq_error(struct ata_link *link); extern void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset, ata_reset_fn_t softreset, ata_reset_fn_t hardreset, @@ -1381,6 +1382,21 @@ static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host) return *(struct ata_port **)&host->hostdata[0]; } +static inline int ata_check_ready(u8 status) +{ + /* Some controllers report 0x77 or 0x7f during intermediate + * not-ready stages. + */ + if (status == 0x77 || status == 0x7f) + return 0; + + /* 0xff indicates either no device or device not ready */ + if (status == 0xff) + return -ENODEV; + + return !(status & ATA_BUSY); +} + /************************************************************************** * PMP - drivers/ata/libata-pmp.c diff --git a/include/net/ip.h b/include/net/ip.h index 6d7bcd5e62d..3b40bc2234b 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -210,7 +210,7 @@ int ip_dont_fragment(struct sock *sk, struct dst_entry *dst) { return (inet_sk(sk)->pmtudisc == IP_PMTUDISC_DO || (inet_sk(sk)->pmtudisc == IP_PMTUDISC_WANT && - !(dst_metric(dst, RTAX_LOCK)&(1<<RTAX_MTU)))); + !(dst_metric_locked(dst, RTAX_MTU)))); } extern void __ip_select_ident(struct iphdr *iph, struct dst_entry *dst, int more); diff --git a/include/net/xfrm.h b/include/net/xfrm.h index d1350bcccb0..2933d7474a7 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -648,14 +648,46 @@ extern void xfrm_audit_state_notfound(struct sk_buff *skb, u16 family, extern void xfrm_audit_state_icvfail(struct xfrm_state *x, struct sk_buff *skb, u8 proto); #else -#define xfrm_audit_policy_add(x, r, a, se, s) do { ; } while (0) -#define xfrm_audit_policy_delete(x, r, a, se, s) do { ; } while (0) -#define xfrm_audit_state_add(x, r, a, se, s) do { ; } while (0) -#define xfrm_audit_state_delete(x, r, a, se, s) do { ; } while (0) -#define xfrm_audit_state_replay_overflow(x, s) do { ; } while (0) -#define xfrm_audit_state_notfound_simple(s, f) do { ; } while (0) -#define xfrm_audit_state_notfound(s, f, sp, sq) do { ; } while (0) -#define xfrm_audit_state_icvfail(x, s, p) do { ; } while (0) + +static inline void xfrm_audit_policy_add(struct xfrm_policy *xp, int result, + u32 auid, u32 ses, u32 secid) +{ +} + +static inline void xfrm_audit_policy_delete(struct xfrm_policy *xp, int result, + u32 auid, u32 ses, u32 secid) +{ +} + +static inline void xfrm_audit_state_add(struct xfrm_state *x, int result, + u32 auid, u32 ses, u32 secid) +{ +} + +static inline void xfrm_audit_state_delete(struct xfrm_state *x, int result, + u32 auid, u32 ses, u32 secid) +{ +} + +static inline void xfrm_audit_state_replay_overflow(struct xfrm_state *x, + struct sk_buff *skb) +{ +} + +static inline void xfrm_audit_state_notfound_simple(struct sk_buff *skb, + u16 family) +{ +} + +static inline void xfrm_audit_state_notfound(struct sk_buff *skb, u16 family, + __be32 net_spi, __be32 net_seq) +{ +} + +static inline void xfrm_audit_state_icvfail(struct xfrm_state *x, + struct sk_buff *skb, u8 proto) +{ +} #endif /* CONFIG_AUDITSYSCALL */ static inline void xfrm_pol_hold(struct xfrm_policy *policy) diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 8da627d3380..86ea9e34e32 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -1031,11 +1031,9 @@ int current_cpuset_is_being_rebound(void) return task_cs(current) == cpuset_being_rebound; } -static int update_relax_domain_level(struct cpuset *cs, char *buf) +static int update_relax_domain_level(struct cpuset *cs, s64 val) { - int val = simple_strtol(buf, NULL, 10); - - if (val < 0) + if ((int)val < 0) val = -1; if (val != cs->relax_domain_level) { @@ -1280,9 +1278,6 @@ static ssize_t cpuset_common_file_write(struct cgroup *cont, case FILE_MEMLIST: retval = update_nodemask(cs, buffer); break; - case FILE_SCHED_RELAX_DOMAIN_LEVEL: - retval = update_relax_domain_level(cs, buffer); - break; default: retval = -EINVAL; goto out2; @@ -1348,6 +1343,30 @@ static int cpuset_write_u64(struct cgroup *cgrp, struct cftype *cft, u64 val) return retval; } +static int cpuset_write_s64(struct cgroup *cgrp, struct cftype *cft, s64 val) +{ + int retval = 0; + struct cpuset *cs = cgroup_cs(cgrp); + cpuset_filetype_t type = cft->private; + + cgroup_lock(); + + if (cgroup_is_removed(cgrp)) { + cgroup_unlock(); + return -ENODEV; + } + switch (type) { + case FILE_SCHED_RELAX_DOMAIN_LEVEL: + retval = update_relax_domain_level(cs, val); + break; + default: + retval = -EINVAL; + break; + } + cgroup_unlock(); + return retval; +} + /* * These ascii lists should be read in a single call, by using a user * buffer large enough to hold the entire map. If read in smaller @@ -1406,9 +1425,6 @@ static ssize_t cpuset_common_file_read(struct cgroup *cont, case FILE_MEMLIST: s += cpuset_sprintf_memlist(s, cs); break; - case FILE_SCHED_RELAX_DOMAIN_LEVEL: - s += sprintf(s, "%d", cs->relax_domain_level); - break; default: retval = -EINVAL; goto out; @@ -1449,6 +1465,18 @@ static u64 cpuset_read_u64(struct cgroup *cont, struct cftype *cft) } } +static s64 cpuset_read_s64(struct cgroup *cont, struct cftype *cft) +{ + struct cpuset *cs = cgroup_cs(cont); + cpuset_filetype_t type = cft->private; + switch (type) { + case FILE_SCHED_RELAX_DOMAIN_LEVEL: + return cs->relax_domain_level; + default: + BUG(); + } +} + /* * for the common functions, 'private' gives the type of file @@ -1499,8 +1527,8 @@ static struct cftype files[] = { { .name = "sched_relax_domain_level", - .read_u64 = cpuset_read_u64, - .write_u64 = cpuset_write_u64, + .read_s64 = cpuset_read_s64, + .write_s64 = cpuset_write_s64, .private = FILE_SCHED_RELAX_DOMAIN_LEVEL, }, diff --git a/kernel/relay.c b/kernel/relay.c index 7de644cdec4..bc24dcdc570 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -1191,7 +1191,7 @@ static ssize_t relay_file_splice_read(struct file *in, ret = 0; spliced = 0; - while (len && !spliced) { + while (len) { ret = subbuf_splice_actor(in, ppos, pipe, len, flags, &nonpad_ret); if (ret < 0) break; diff --git a/mm/filemap.c b/mm/filemap.c index 239d36163bb..2dead9adf8b 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -1655,7 +1655,7 @@ int should_remove_suid(struct dentry *dentry) } EXPORT_SYMBOL(should_remove_suid); -int __remove_suid(struct dentry *dentry, int kill) +static int __remove_suid(struct dentry *dentry, int kill) { struct iattr newattrs; diff --git a/mm/memory.c b/mm/memory.c index bbab1e37055..48c122d42ed 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -969,7 +969,7 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address, goto no_page_table; pmd = pmd_offset(pud, address); - if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd))) + if (pmd_none(*pmd)) goto no_page_table; if (pmd_huge(*pmd)) { @@ -978,6 +978,9 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address, goto out; } + if (unlikely(pmd_bad(*pmd))) + goto no_page_table; + ptep = pte_offset_map_lock(mm, pmd, address, &ptl); if (!ptep) goto out; diff --git a/mm/slub.c b/mm/slub.c index d379b782fc8..a505a828ef4 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -3762,7 +3762,7 @@ static int any_slab_objects(struct kmem_cache *s) if (!n) continue; - if (atomic_read(&n->total_objects)) + if (atomic_long_read(&n->total_objects)) return 1; } return 0; diff --git a/net/atm/br2684.c b/net/atm/br2684.c index 1b228065e74..9d52ebfc196 100644 --- a/net/atm/br2684.c +++ b/net/atm/br2684.c @@ -346,9 +346,9 @@ static void br2684_push(struct atm_vcc *atmvcc, struct sk_buff *skb) /* skb==NULL means VCC is being destroyed */ br2684_close_vcc(brvcc); if (list_empty(&brdev->brvccs)) { - read_lock(&devs_lock); + write_lock_irq(&devs_lock); list_del(&brdev->br2684_devs); - read_unlock(&devs_lock); + write_unlock_irq(&devs_lock); unregister_netdev(net_dev); free_netdev(net_dev); } diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c index 77a981a1ee5..c2397f503b0 100644 --- a/net/bridge/br_if.c +++ b/net/bridge/br_if.c @@ -273,15 +273,13 @@ int br_add_bridge(const char *name) rtnl_lock(); if (strchr(dev->name, '%')) { ret = dev_alloc_name(dev, dev->name); - if (ret < 0) { - free_netdev(dev); - goto out; - } + if (ret < 0) + goto out_free; } ret = register_netdevice(dev); if (ret) - goto out; + goto out_free; ret = br_sysfs_addbr(dev); if (ret) @@ -289,6 +287,10 @@ int br_add_bridge(const char *name) out: rtnl_unlock(); return ret; + +out_free: + free_netdev(dev); + goto out; } int br_del_bridge(const char *name) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 4fe605fa6f8..5c459f2b798 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -200,7 +200,9 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, goto nodata; /* - * See comment in sk_buff definition, just before the 'tail' member + * Only clear those fields we need to clear, not those that we will + * actually initialise below. Hence, don't put any more fields after + * the tail pointer in struct sk_buff! */ memset(skb, 0, offsetof(struct sk_buff, tail)); skb->truesize = size + sizeof(struct sk_buff); diff --git a/net/dccp/feat.c b/net/dccp/feat.c index 4a4f6ce4498..933a0ecf8d4 100644 --- a/net/dccp/feat.c +++ b/net/dccp/feat.c @@ -32,7 +32,7 @@ int dccp_feat_change(struct dccp_minisock *dmsk, u8 type, u8 feature, if (len > 3) { DCCP_WARN("invalid length %d\n", len); - return 1; + return -EINVAL; } /* XXX add further sanity checks */ diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c index 2f665a51647..f50e88bf266 100644 --- a/net/decnet/dn_route.c +++ b/net/decnet/dn_route.c @@ -235,14 +235,14 @@ static void dn_dst_update_pmtu(struct dst_entry *dst, u32 mtu) else min_mtu -= 21; - if (dst->metrics[RTAX_MTU-1] > mtu && mtu >= min_mtu) { + if (dst_metric(dst, RTAX_MTU) > mtu && mtu >= min_mtu) { if (!(dst_metric_locked(dst, RTAX_MTU))) { dst->metrics[RTAX_MTU-1] = mtu; dst_set_expires(dst, dn_rt_mtu_expires); } if (!(dst_metric_locked(dst, RTAX_ADVMSS))) { u32 mss = mtu - DN_MAX_NSP_DATA_HEADER; - if (dst->metrics[RTAX_ADVMSS-1] > mss) + if (dst_metric(dst, RTAX_ADVMSS) > mss) dst->metrics[RTAX_ADVMSS-1] = mss; } } @@ -805,12 +805,12 @@ static int dn_rt_set_next_hop(struct dn_route *rt, struct dn_fib_res *res) rt->u.dst.neighbour = n; } - if (rt->u.dst.metrics[RTAX_MTU-1] == 0 || - rt->u.dst.metrics[RTAX_MTU-1] > rt->u.dst.dev->mtu) + if (dst_metric(&rt->u.dst, RTAX_MTU) == 0 || + dst_metric(&rt->u.dst, RTAX_MTU) > rt->u.dst.dev->mtu) rt->u.dst.metrics[RTAX_MTU-1] = rt->u.dst.dev->mtu; mss = dn_mss_from_pmtu(dev, dst_mtu(&rt->u.dst)); - if (rt->u.dst.metrics[RTAX_ADVMSS-1] == 0 || - rt->u.dst.metrics[RTAX_ADVMSS-1] > mss) + if (dst_metric(&rt->u.dst, RTAX_ADVMSS) == 0 || + dst_metric(&rt->u.dst, RTAX_ADVMSS) > mss) rt->u.dst.metrics[RTAX_ADVMSS-1] = mss; return 0; } diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 5e3685c5c40..92f90ae46f4 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1468,14 +1468,14 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, /* BSD 4.2 compatibility hack :-( */ if (mtu == 0 && - old_mtu >= rth->u.dst.metrics[RTAX_MTU-1] && + old_mtu >= dst_metric(&rth->u.dst, RTAX_MTU) && old_mtu >= 68 + (iph->ihl << 2)) old_mtu -= iph->ihl << 2; mtu = guess_mtu(old_mtu); } - if (mtu <= rth->u.dst.metrics[RTAX_MTU-1]) { - if (mtu < rth->u.dst.metrics[RTAX_MTU-1]) { + if (mtu <= dst_metric(&rth->u.dst, RTAX_MTU)) { + if (mtu < dst_metric(&rth->u.dst, RTAX_MTU)) { dst_confirm(&rth->u.dst); if (mtu < ip_rt_min_pmtu) { mtu = ip_rt_min_pmtu; @@ -1497,7 +1497,7 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, static void ip_rt_update_pmtu(struct dst_entry *dst, u32 mtu) { - if (dst->metrics[RTAX_MTU-1] > mtu && mtu >= 68 && + if (dst_metric(dst, RTAX_MTU) > mtu && mtu >= 68 && !(dst_metric_locked(dst, RTAX_MTU))) { if (mtu < ip_rt_min_pmtu) { mtu = ip_rt_min_pmtu; @@ -1613,7 +1613,7 @@ static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag) sizeof(rt->u.dst.metrics)); if (fi->fib_mtu == 0) { rt->u.dst.metrics[RTAX_MTU-1] = rt->u.dst.dev->mtu; - if (rt->u.dst.metrics[RTAX_LOCK-1] & (1 << RTAX_MTU) && + if (dst_metric_locked(&rt->u.dst, RTAX_MTU) && rt->rt_gateway != rt->rt_dst && rt->u.dst.dev->mtu > 576) rt->u.dst.metrics[RTAX_MTU-1] = 576; @@ -1624,14 +1624,14 @@ static void rt_set_nexthop(struct rtable *rt, struct fib_result *res, u32 itag) } else rt->u.dst.metrics[RTAX_MTU-1]= rt->u.dst.dev->mtu; - if (rt->u.dst.metrics[RTAX_HOPLIMIT-1] == 0) + if (dst_metric(&rt->u.dst, RTAX_HOPLIMIT) == 0) rt->u.dst.metrics[RTAX_HOPLIMIT-1] = sysctl_ip_default_ttl; - if (rt->u.dst.metrics[RTAX_MTU-1] > IP_MAX_MTU) + if (dst_metric(&rt->u.dst, RTAX_MTU) > IP_MAX_MTU) rt->u.dst.metrics[RTAX_MTU-1] = IP_MAX_MTU; - if (rt->u.dst.metrics[RTAX_ADVMSS-1] == 0) + if (dst_metric(&rt->u.dst, RTAX_ADVMSS) == 0) rt->u.dst.metrics[RTAX_ADVMSS-1] = max_t(unsigned int, rt->u.dst.dev->mtu - 40, ip_rt_min_advmss); - if (rt->u.dst.metrics[RTAX_ADVMSS-1] > 65535 - 40) + if (dst_metric(&rt->u.dst, RTAX_ADVMSS) > 65535 - 40) rt->u.dst.metrics[RTAX_ADVMSS-1] = 65535 - 40; #ifdef CONFIG_NET_CLS_ROUTE diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index eda4f4a233f..8ac15a604e0 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -66,6 +66,7 @@ #include <linux/mm.h> #include <linux/module.h> #include <linux/sysctl.h> +#include <net/dst.h> #include <net/tcp.h> #include <net/inet_common.h> #include <linux/ipsec.h> @@ -605,7 +606,7 @@ static u32 tcp_rto_min(struct sock *sk) u32 rto_min = TCP_RTO_MIN; if (dst && dst_metric_locked(dst, RTAX_RTO_MIN)) - rto_min = dst->metrics[RTAX_RTO_MIN - 1]; + rto_min = dst_metric(dst, RTAX_RTO_MIN); return rto_min; } @@ -769,7 +770,7 @@ void tcp_update_metrics(struct sock *sk) dst->metrics[RTAX_RTTVAR - 1] = m; else dst->metrics[RTAX_RTTVAR-1] -= - (dst->metrics[RTAX_RTTVAR-1] - m)>>2; + (dst_metric(dst, RTAX_RTTVAR) - m)>>2; } if (tp->snd_ssthresh >= 0xFFFF) { @@ -788,21 +789,21 @@ void tcp_update_metrics(struct sock *sk) dst->metrics[RTAX_SSTHRESH-1] = max(tp->snd_cwnd >> 1, tp->snd_ssthresh); if (!dst_metric_locked(dst, RTAX_CWND)) - dst->metrics[RTAX_CWND-1] = (dst->metrics[RTAX_CWND-1] + tp->snd_cwnd) >> 1; + dst->metrics[RTAX_CWND-1] = (dst_metric(dst, RTAX_CWND) + tp->snd_cwnd) >> 1; } else { /* Else slow start did not finish, cwnd is non-sense, ssthresh may be also invalid. */ if (!dst_metric_locked(dst, RTAX_CWND)) - dst->metrics[RTAX_CWND-1] = (dst->metrics[RTAX_CWND-1] + tp->snd_ssthresh) >> 1; - if (dst->metrics[RTAX_SSTHRESH-1] && + dst->metrics[RTAX_CWND-1] = (dst_metric(dst, RTAX_CWND) + tp->snd_ssthresh) >> 1; + if (dst_metric(dst, RTAX_SSTHRESH) && !dst_metric_locked(dst, RTAX_SSTHRESH) && - tp->snd_ssthresh > dst->metrics[RTAX_SSTHRESH-1]) + tp->snd_ssthresh > dst_metric(dst, RTAX_SSTHRESH)) dst->metrics[RTAX_SSTHRESH-1] = tp->snd_ssthresh; } if (!dst_metric_locked(dst, RTAX_REORDERING)) { - if (dst->metrics[RTAX_REORDERING-1] < tp->reordering && + if (dst_metric(dst, RTAX_REORDERING) < tp->reordering && tp->reordering != sysctl_tcp_reordering) dst->metrics[RTAX_REORDERING-1] = tp->reordering; } diff --git a/net/ipv6/route.c b/net/ipv6/route.c index a493ad9b891..12bba088034 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1243,11 +1243,11 @@ install_route: } } - if (rt->u.dst.metrics[RTAX_HOPLIMIT-1] == 0) + if (dst_metric(&rt->u.dst, RTAX_HOPLIMIT) == 0) rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1; - if (!rt->u.dst.metrics[RTAX_MTU-1]) + if (!dst_metric(&rt->u.dst, RTAX_MTU)) rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(dev); - if (!rt->u.dst.metrics[RTAX_ADVMSS-1]) + if (!dst_metric(&rt->u.dst, RTAX_ADVMSS)) rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(net, dst_mtu(&rt->u.dst)); rt->u.dst.dev = dev; rt->rt6i_idev = idev; diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 9ad4e3631b6..915afadb060 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -1766,6 +1766,7 @@ fail_wep: fail_rate: ieee80211_debugfs_remove_netdev(IEEE80211_DEV_TO_SUB_IF(local->mdev)); unregister_netdevice(local->mdev); + local->mdev = NULL; fail_dev: rtnl_unlock(); sta_info_stop(local); @@ -1773,8 +1774,10 @@ fail_sta_info: debugfs_hw_del(local); destroy_workqueue(local->hw.workqueue); fail_workqueue: - ieee80211_if_free(local->mdev); - local->mdev = NULL; + if (local->mdev != NULL) { + ieee80211_if_free(local->mdev); + local->mdev = NULL; + } fail_mdev_alloc: wiphy_unregister(local->hw.wiphy); return result; diff --git a/net/mac80211/rc80211_pid_debugfs.c b/net/mac80211/rc80211_pid_debugfs.c index ae75d417873..ff5c380f3c1 100644 --- a/net/mac80211/rc80211_pid_debugfs.c +++ b/net/mac80211/rc80211_pid_debugfs.c @@ -85,7 +85,7 @@ static int rate_control_pid_events_open(struct inode *inode, struct file *file) struct rc_pid_sta_info *sinfo = inode->i_private; struct rc_pid_event_buffer *events = &sinfo->events; struct rc_pid_events_file_info *file_info; - unsigned int status; + unsigned long status; /* Allocate a state struct */ file_info = kmalloc(sizeof(*file_info), GFP_KERNEL); @@ -135,7 +135,7 @@ static ssize_t rate_control_pid_events_read(struct file *file, char __user *buf, char pb[RC_PID_PRINT_BUF_SIZE]; int ret; int p; - unsigned int status; + unsigned long status; /* Check if there is something to read. */ if (events->next_entry == file_info->next_entry) { diff --git a/net/sched/act_simple.c b/net/sched/act_simple.c index 64b2d136c78..1d421d059ca 100644 --- a/net/sched/act_simple.c +++ b/net/sched/act_simple.c @@ -6,7 +6,7 @@ * as published by the Free Software Foundation; either version * 2 of the License, or (at your option) any later version. * - * Authors: Jamal Hadi Salim (2005) + * Authors: Jamal Hadi Salim (2005-8) * */ @@ -34,6 +34,7 @@ static struct tcf_hashinfo simp_hash_info = { .lock = &simp_lock, }; +#define SIMP_MAX_DATA 32 static int tcf_simp(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res) { struct tcf_defact *d = a->priv; @@ -69,23 +70,28 @@ static int tcf_simp_release(struct tcf_defact *d, int bind) return ret; } -static int alloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata) +static int alloc_defdata(struct tcf_defact *d, char *defdata) { - d->tcfd_defdata = kmemdup(defdata, datalen, GFP_KERNEL); + d->tcfd_defdata = kstrndup(defdata, SIMP_MAX_DATA, GFP_KERNEL); if (unlikely(!d->tcfd_defdata)) return -ENOMEM; - d->tcfd_datalen = datalen; + return 0; } -static int realloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata) +static void reset_policy(struct tcf_defact *d, char *defdata, + struct tc_defact *p) { - kfree(d->tcfd_defdata); - return alloc_defdata(d, datalen, defdata); + spin_lock_bh(&d->tcf_lock); + d->tcf_action = p->action; + memset(d->tcfd_defdata, 0, SIMP_MAX_DATA); + strlcpy(d->tcfd_defdata, defdata, SIMP_MAX_DATA); + spin_unlock_bh(&d->tcf_lock); } static const struct nla_policy simple_policy[TCA_DEF_MAX + 1] = { [TCA_DEF_PARMS] = { .len = sizeof(struct tc_defact) }, + [TCA_DEF_DATA] = { .type = NLA_STRING, .len = SIMP_MAX_DATA }, }; static int tcf_simp_init(struct nlattr *nla, struct nlattr *est, @@ -95,28 +101,24 @@ static int tcf_simp_init(struct nlattr *nla, struct nlattr *est, struct tc_defact *parm; struct tcf_defact *d; struct tcf_common *pc; - void *defdata; - u32 datalen = 0; + char *defdata; int ret = 0, err; if (nla == NULL) return -EINVAL; - err = nla_parse_nested(tb, TCA_DEF_MAX, nla, NULL); + err = nla_parse_nested(tb, TCA_DEF_MAX, nla, simple_policy); if (err < 0) return err; if (tb[TCA_DEF_PARMS] == NULL) return -EINVAL; - parm = nla_data(tb[TCA_DEF_PARMS]); - defdata = nla_data(tb[TCA_DEF_DATA]); - if (defdata == NULL) + if (tb[TCA_DEF_DATA] == NULL) return -EINVAL; - datalen = nla_len(tb[TCA_DEF_DATA]); - if (datalen == 0) - return -EINVAL; + parm = nla_data(tb[TCA_DEF_PARMS]); + defdata = nla_data(tb[TCA_DEF_DATA]); pc = tcf_hash_check(parm->index, a, bind, &simp_hash_info); if (!pc) { @@ -126,11 +128,12 @@ static int tcf_simp_init(struct nlattr *nla, struct nlattr *est, return -ENOMEM; d = to_defact(pc); - ret = alloc_defdata(d, datalen, defdata); + ret = alloc_defdata(d, defdata); if (ret < 0) { kfree(pc); return ret; } + d->tcf_action = parm->action; ret = ACT_P_CREATED; } else { d = to_defact(pc); @@ -138,13 +141,9 @@ static int tcf_simp_init(struct nlattr *nla, struct nlattr *est, tcf_simp_release(d, bind); return -EEXIST; } - realloc_defdata(d, datalen, defdata); + reset_policy(d, defdata, parm); } - spin_lock_bh(&d->tcf_lock); - d->tcf_action = parm->action; - spin_unlock_bh(&d->tcf_lock); - if (ret == ACT_P_CREATED) tcf_hash_insert(pc, &simp_hash_info); return ret; @@ -172,7 +171,7 @@ static inline int tcf_simp_dump(struct sk_buff *skb, struct tc_action *a, opt.bindcnt = d->tcf_bindcnt - bind; opt.action = d->tcf_action; NLA_PUT(skb, TCA_DEF_PARMS, sizeof(opt), &opt); - NLA_PUT(skb, TCA_DEF_DATA, d->tcfd_datalen, d->tcfd_defdata); + NLA_PUT_STRING(skb, TCA_DEF_DATA, d->tcfd_defdata); t.install = jiffies_to_clock_t(jiffies - d->tcf_tm.install); t.lastuse = jiffies_to_clock_t(jiffies - d->tcf_tm.lastuse); t.expires = jiffies_to_clock_t(d->tcf_tm.expires); diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 66148cc4759..5bc1ed49018 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -1197,12 +1197,16 @@ static inline int htb_parent_last_child(struct htb_class *cl) return 1; } -static void htb_parent_to_leaf(struct htb_class *cl, struct Qdisc *new_q) +static void htb_parent_to_leaf(struct htb_sched *q, struct htb_class *cl, + struct Qdisc *new_q) { struct htb_class *parent = cl->parent; BUG_TRAP(!cl->level && cl->un.leaf.q && !cl->prio_activity); + if (parent->cmode != HTB_CAN_SEND) + htb_safe_rb_erase(&parent->pq_node, q->wait_pq + parent->level); + parent->level = 0; memset(&parent->un.inner, 0, sizeof(parent->un.inner)); INIT_LIST_HEAD(&parent->un.leaf.drop_list); @@ -1300,7 +1304,7 @@ static int htb_delete(struct Qdisc *sch, unsigned long arg) htb_deactivate(q, cl); if (last_child) - htb_parent_to_leaf(cl, new_q); + htb_parent_to_leaf(q, cl, new_q); if (--cl->refcnt == 0) htb_destroy_class(sch, cl); diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index d74c2d26953..01c7e311b90 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -18,7 +18,6 @@ #include <linux/mm.h> #include <linux/interrupt.h> #include <linux/module.h> -#include <linux/sched.h> #include <linux/sunrpc/types.h> #include <linux/sunrpc/xdr.h> diff --git a/sound/oss/kahlua.c b/sound/oss/kahlua.c index dfe670f12e6..eb9bc365530 100644 --- a/sound/oss/kahlua.c +++ b/sound/oss/kahlua.c @@ -67,7 +67,7 @@ static int __devinit probe_one(struct pci_dev *pdev, const struct pci_device_id return 1; mem = ioremap(base, 128); - if(mem == 0UL) + if (!mem) return 1; map = readw(mem + 0x18); /* Read the SMI enables */ iounmap(mem); |