From 6c2e94033df5ca11149e52dd179b8dde3172e9bf Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 7 Nov 2008 12:33:49 +0100 Subject: x86: apic honour irq affinity which was set in early boot setup_ioapic_dest() is called after the non boot cpus have been brought up. It sets the irq affinity of all already configured interrupts to all cpus and ignores affinity settings which were done by the early bootup code. If the IRQ_NO_BALANCING or IRQ_AFFINITY_SET flags are set then use the affinity mask from the irq descriptor and not TARGET_CPUS. Signed-off-by: Thomas Gleixner Signed-off-by: Ingo Molnar --- arch/x86/kernel/io_apic.c | 26 +++++++++++++++++++++----- 1 file changed, 21 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c index 7a3f2028e2eb..988ee89467d3 100644 --- a/arch/x86/kernel/io_apic.c +++ b/arch/x86/kernel/io_apic.c @@ -3761,7 +3761,9 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity) void __init setup_ioapic_dest(void) { int pin, ioapic, irq, irq_entry; + struct irq_desc *desc; struct irq_cfg *cfg; + cpumask_t mask; if (skip_ioapic_setup == 1) return; @@ -3778,16 +3780,30 @@ void __init setup_ioapic_dest(void) * cpu is online. */ cfg = irq_cfg(irq); - if (!cfg->vector) + if (!cfg->vector) { setup_IO_APIC_irq(ioapic, pin, irq, irq_trigger(irq_entry), irq_polarity(irq_entry)); + continue; + + } + + /* + * Honour affinities which have been set in early boot + */ + desc = irq_to_desc(irq); + if (desc->status & + (IRQ_NO_BALANCING | IRQ_AFFINITY_SET)) + mask = desc->affinity; + else + mask = TARGET_CPUS; + #ifdef CONFIG_INTR_REMAP - else if (intr_remapping_enabled) - set_ir_ioapic_affinity_irq(irq, TARGET_CPUS); -#endif + if (intr_remapping_enabled) + set_ir_ioapic_affinity_irq(irq, mask); else - set_ioapic_affinity_irq(irq, TARGET_CPUS); +#endif + set_ioapic_affinity_irq(irq, mask); } } -- cgit v1.2.3 From a266d9f1253a38ec2d5655ebcd6846298b0554f4 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 21 Nov 2008 14:49:25 +0100 Subject: [CPUFREQ] powernow-k8: ignore out-of-range PstateStatus value A workaround for AMD CPU family 11h erratum 311 might cause that the P-state Status Register shows a "current P-state" which is larger than the "current P-state limit" in P-state Current Limit Register. For the wrong P-state value there is no ACPI _PSS object defined and powernow-k8/cpufreq can't determine the proper CPU frequency for that state. As a consequence this can cause a panic during boot (potentially with all recent kernel versions -- at least I have reproduced it with various 2.6.27 kernels and with the current .28 series), as an example: powernow-k8: Found 1 AMD Turion(tm)X2 Ultra DualCore Mobile ZM-82 processors (2 \ ) powernow-k8: 0 : pstate 0 (2200 MHz) powernow-k8: 1 : pstate 1 (1100 MHz) powernow-k8: 2 : pstate 2 (600 MHz) BUG: unable to handle kernel paging request at ffff88086e7528b8 IP: [] cpufreq_stats_update+0x4a/0x5f PGD 202063 PUD 0 Oops: 0002 [#1] SMP last sysfs file: CPU 1 Modules linked in: Pid: 1, comm: swapper Not tainted 2.6.28-rc3-dirty #16 RIP: 0010:[] [] cpufreq_stats_update+0x4a/0\ f Synaptics claims to have extended capabilities, but I'm not able to read them.<6\ 6 RAX: 0000000000000000 RBX: 0000000000000001 RCX: ffff88006e7528c0 RDX: 00000000ffffffff RSI: ffff88006e54af00 RDI: ffffffff808f056c RBP: 00000000fffee697 R08: 0000000000000003 R09: ffff88006e73f080 R10: 0000000000000001 R11: 00000000002191c0 R12: ffff88006fb83c10 R13: 00000000ffffffff R14: 0000000000000001 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffff88006fb50740(0000) knlGS:0000000000000000 Unable to initialize Synaptics hardware. CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: ffff88086e7528b8 CR3: 0000000000201000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process swapper (pid: 1, threadinfo ffff88006fb82000, task ffff88006fb816d0) Stack: ffff88006e74da50 0000000000000000 ffff88006e54af00 ffffffff804863c7 ffff88006e74da50 0000000000000000 00000000ffffffff 0000000000000000 ffff88006fb83c10 ffffffff8024b46c ffffffff808f0560 ffff88006fb83c10 Call Trace: [] ? cpufreq_stat_notifier_trans+0x51/0x83 [] ? notifier_call_chain+0x29/0x4c [] ? __srcu_notifier_call_chain+0x46/0x61 [] ? cpufreq_notify_transition+0x93/0xa9 [] ? powernowk8_target+0x1e8/0x5f3 [] ? cpufreq_governor_performance+0x1b/0x20 [] ? __cpufreq_governor+0x71/0xa8 [] ? __cpufreq_set_policy+0x101/0x13e [] ? cpufreq_add_dev+0x3f0/0x4cd [] ? handle_update+0x0/0x8 [] ? sysdev_driver_register+0xb6/0x10d [] ? powernowk8_init+0x0/0x7e [] ? cpufreq_register_driver+0x8f/0x140 [] ? _stext+0x56/0x14f [] ? proc_register+0x122/0x17d [] ? create_proc_entry+0x73/0x8a [] ? register_irq_proc+0x92/0xaa [] ? init_irq_proc+0x57/0x69 [] ? kernel_init+0x116/0x169 [] ? child_rip+0xa/0x11 [] ? kernel_init+0x0/0x169 [] ? child_rip+0x0/0x11 Code: 05 c5 83 36 00 48 c7 c2 48 5d 86 80 48 8b 04 d8 48 8b 40 08 48 8b 34 02 48\ RIP [] cpufreq_stats_update+0x4a/0x5f RSP CR2: ffff88086e7528b8 ---[ end trace 0678bac75e67a2f7 ]--- Kernel panic - not syncing: Attempted to kill init! In short, aftereffect of the wrong P-state is that cpufreq_stats_update() uses "-1" as index for some array in cpufreq_stats_update (unsigned int cpu) { ... if (stat->time_in_state) stat->time_in_state[stat->last_index] = cputime64_add(stat->time_in_state[stat->last_index], cputime_sub(cur_time, stat->last_time)); ... } Fortunately, the wrong P-state value is returned only if the core is in P-state 0. This fix solves the problem by detecting the out-of-range P-state, ignoring it, and using "0" instead. Cc: Mark Langsdorf Signed-off-by: Andreas Herrmann Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 18 +++++++++++++++--- arch/x86/kernel/cpu/cpufreq/powernow-k8.h | 17 ++++++++++++++++- 2 files changed, 31 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index d3dcd58b87cd..7f05f44b97e9 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -115,9 +115,20 @@ static int query_current_values_with_pending_wait(struct powernow_k8_data *data) u32 i = 0; if (cpu_family == CPU_HW_PSTATE) { - rdmsr(MSR_PSTATE_STATUS, lo, hi); - i = lo & HW_PSTATE_MASK; - data->currpstate = i; + if (data->currpstate == HW_PSTATE_INVALID) { + /* read (initial) hw pstate if not yet set */ + rdmsr(MSR_PSTATE_STATUS, lo, hi); + i = lo & HW_PSTATE_MASK; + + /* + * a workaround for family 11h erratum 311 might cause + * an "out-of-range Pstate if the core is in Pstate-0 + */ + if (i >= data->numps) + data->currpstate = HW_PSTATE_0; + else + data->currpstate = i; + } return 0; } do { @@ -1121,6 +1132,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) } data->cpu = pol->cpu; + data->currpstate = HW_PSTATE_INVALID; if (powernow_k8_cpu_init_acpi(data)) { /* diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h index ab48cfed4d96..65cfb5d7f77f 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h @@ -5,6 +5,19 @@ * http://www.gnu.org/licenses/gpl.html */ + +enum pstate { + HW_PSTATE_INVALID = 0xff, + HW_PSTATE_0 = 0, + HW_PSTATE_1 = 1, + HW_PSTATE_2 = 2, + HW_PSTATE_3 = 3, + HW_PSTATE_4 = 4, + HW_PSTATE_5 = 5, + HW_PSTATE_6 = 6, + HW_PSTATE_7 = 7, +}; + struct powernow_k8_data { unsigned int cpu; @@ -23,7 +36,9 @@ struct powernow_k8_data { u32 exttype; /* extended interface = 1 */ /* keep track of the current fid / vid or pstate */ - u32 currvid, currfid, currpstate; + u32 currvid; + u32 currfid; + enum pstate currpstate; /* the powernow_table includes all frequency and vid/fid pairings: * fid are the lower 8 bits of the index, vid are the upper 8 bits. -- cgit v1.2.3 From 2236d252e001ea57d53cec1954f680e503f3b8bc Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 22 Nov 2008 17:37:34 +0000 Subject: enable_IR_x2apic() needs to be __init calls __init, called only from __init Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/kernel/apic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/apic.c b/arch/x86/kernel/apic.c index 04a7f960bbc0..16f94879b525 100644 --- a/arch/x86/kernel/apic.c +++ b/arch/x86/kernel/apic.c @@ -1315,7 +1315,7 @@ void enable_x2apic(void) } } -void enable_IR_x2apic(void) +void __init enable_IR_x2apic(void) { #ifdef CONFIG_INTR_REMAP int ret; -- cgit v1.2.3 From 23a14b9e9db49ed5f7683857557c26c874d4abb6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 22 Nov 2008 17:37:44 +0000 Subject: kvm_setup_secondary_clock() is cpuinit Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/kernel/kvmclock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c index 1c9cc431ea4f..e169ae9b6a62 100644 --- a/arch/x86/kernel/kvmclock.c +++ b/arch/x86/kernel/kvmclock.c @@ -128,7 +128,7 @@ static int kvm_register_clock(char *txt) } #ifdef CONFIG_X86_LOCAL_APIC -static void __devinit kvm_setup_secondary_clock(void) +static void __cpuinit kvm_setup_secondary_clock(void) { /* * Now that the first cpu already had this clocksource initialized, -- cgit v1.2.3 From 37af46efa5413c6f4c25d9a24b4c43f2cc718eed Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 22 Nov 2008 17:37:54 +0000 Subject: xen_setup_vcpu_info_placement() is not init on x86 ... so get xen-ops.h in agreement with xen/smp.c Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/xen/xen-ops.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index d7422dc2a55c..9e1afae8461f 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -49,7 +49,7 @@ bool xen_vcpu_stolen(int vcpu); void xen_mark_init_mm_pinned(void); -void __init xen_setup_vcpu_info_placement(void); +void xen_setup_vcpu_info_placement(void); #ifdef CONFIG_SMP void xen_smp_init(void); -- cgit v1.2.3 From df6b07949b6cab9d119363d02ef63379160f6c82 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 22 Nov 2008 17:38:04 +0000 Subject: xen_play_dead() is __cpuinit Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/xen/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index d77da613b1d2..acd9b6705e02 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -362,7 +362,7 @@ static void xen_cpu_die(unsigned int cpu) alternatives_smp_switch(0); } -static void xen_play_dead(void) +static void __cpuinit xen_play_dead(void) /* used only with CPU_HOTPLUG */ { play_dead_common(); HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL); -- cgit v1.2.3 From 96b8936a9ed08746e47081458a5eb9e43a751e24 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 25 Nov 2008 08:10:03 +0100 Subject: remove __ARCH_WANT_COMPAT_SYS_PTRACE All architectures now use the generic compat_sys_ptrace, as should every new architecture that needs 32bit compat (if we'll ever get another). Remove the now superflous __ARCH_WANT_COMPAT_SYS_PTRACE define, and also kill a comment about __ARCH_SYS_PTRACE that was added after __ARCH_SYS_PTRACE was already gone. Signed-off-by: Christoph Hellwig Acked-by: David S. Miller Signed-off-by: Linus Torvalds --- arch/Kconfig | 2 -- arch/ia64/include/asm/ptrace.h | 2 -- arch/mips/include/asm/ptrace.h | 4 ---- arch/parisc/include/asm/ptrace.h | 2 -- arch/powerpc/include/asm/ptrace.h | 2 -- arch/s390/include/asm/ptrace.h | 2 -- arch/sparc/include/asm/ptrace_64.h | 2 -- arch/x86/include/asm/ptrace.h | 2 -- include/linux/compat.h | 2 -- kernel/ptrace.c | 4 ++-- 10 files changed, 2 insertions(+), 22 deletions(-) (limited to 'arch/x86') diff --git a/arch/Kconfig b/arch/Kconfig index 8977d99987cb..471e72dbaf8b 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -79,8 +79,6 @@ config HAVE_KRETPROBES # task_pt_regs() in asm/processor.h or asm/ptrace.h # arch_has_single_step() if there is hardware single-step support # arch_has_block_step() if there is hardware block-step support -# arch_ptrace() and not #define __ARCH_SYS_PTRACE -# compat_arch_ptrace() and #define __ARCH_WANT_COMPAT_SYS_PTRACE # asm/syscall.h supplying asm-generic/syscall.h interface # linux/regset.h user_regset interfaces # CORE_DUMP_USE_REGSET #define'd in linux/elf.h diff --git a/arch/ia64/include/asm/ptrace.h b/arch/ia64/include/asm/ptrace.h index 6417c1ecb44e..14055c636adf 100644 --- a/arch/ia64/include/asm/ptrace.h +++ b/arch/ia64/include/asm/ptrace.h @@ -325,8 +325,6 @@ static inline unsigned long user_stack_pointer(struct pt_regs *regs) #define arch_has_block_step() (1) extern void user_enable_block_step(struct task_struct *); -#define __ARCH_WANT_COMPAT_SYS_PTRACE - #endif /* !__KERNEL__ */ /* pt_all_user_regs is used for PTRACE_GETREGS PTRACE_SETREGS */ diff --git a/arch/mips/include/asm/ptrace.h b/arch/mips/include/asm/ptrace.h index 813abd16255d..c2c8bac43307 100644 --- a/arch/mips/include/asm/ptrace.h +++ b/arch/mips/include/asm/ptrace.h @@ -9,10 +9,6 @@ #ifndef _ASM_PTRACE_H #define _ASM_PTRACE_H -#ifdef CONFIG_64BIT -#define __ARCH_WANT_COMPAT_SYS_PTRACE -#endif - /* 0 - 31 are integer registers, 32 - 63 are fp registers. */ #define FPR_BASE 32 #define PC 64 diff --git a/arch/parisc/include/asm/ptrace.h b/arch/parisc/include/asm/ptrace.h index afa5333187b4..302f68dc889c 100644 --- a/arch/parisc/include/asm/ptrace.h +++ b/arch/parisc/include/asm/ptrace.h @@ -47,8 +47,6 @@ struct pt_regs { #define task_regs(task) ((struct pt_regs *) ((char *)(task) + TASK_REGS)) -#define __ARCH_WANT_COMPAT_SYS_PTRACE - struct task_struct; #define arch_has_single_step() 1 void user_disable_single_step(struct task_struct *task); diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h index 280a90cc9894..c9c678fb2538 100644 --- a/arch/powerpc/include/asm/ptrace.h +++ b/arch/powerpc/include/asm/ptrace.h @@ -55,8 +55,6 @@ struct pt_regs { #ifdef __powerpc64__ -#define __ARCH_WANT_COMPAT_SYS_PTRACE - #define STACK_FRAME_OVERHEAD 112 /* size of minimum stack frame */ #define STACK_FRAME_LR_SAVE 2 /* Location of LR in stack frame */ #define STACK_FRAME_REGS_MARKER ASM_CONST(0x7265677368657265) diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h index a7226f8143fb..560ce8561dfd 100644 --- a/arch/s390/include/asm/ptrace.h +++ b/arch/s390/include/asm/ptrace.h @@ -486,8 +486,6 @@ 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 user_stack_pointer(regs)((regs)->gprs[15]) diff --git a/arch/sparc/include/asm/ptrace_64.h b/arch/sparc/include/asm/ptrace_64.h index 3d3e9c161d8b..84e969f06afe 100644 --- a/arch/sparc/include/asm/ptrace_64.h +++ b/arch/sparc/include/asm/ptrace_64.h @@ -142,8 +142,6 @@ struct global_reg_snapshot { }; extern struct global_reg_snapshot global_reg_snapshot[NR_CPUS]; -#define __ARCH_WANT_COMPAT_SYS_PTRACE - #define force_successful_syscall_return() \ do { current_thread_info()->syscall_noerror = 1; \ } while (0) diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index d1531c8480b7..eefb0594b058 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -271,8 +271,6 @@ extern int do_get_thread_area(struct task_struct *p, int idx, extern int do_set_thread_area(struct task_struct *p, int idx, struct user_desc __user *info, int can_allocate); -#define __ARCH_WANT_COMPAT_SYS_PTRACE - #endif /* __KERNEL__ */ #endif /* !__ASSEMBLY__ */ diff --git a/include/linux/compat.h b/include/linux/compat.h index f061a1ea1b74..e88f3ecf38b4 100644 --- a/include/linux/compat.h +++ b/include/linux/compat.h @@ -252,12 +252,10 @@ extern int compat_ptrace_request(struct task_struct *child, compat_long_t request, compat_ulong_t addr, compat_ulong_t data); -#ifdef __ARCH_WANT_COMPAT_SYS_PTRACE extern long compat_arch_ptrace(struct task_struct *child, compat_long_t request, compat_ulong_t addr, compat_ulong_t data); asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid, compat_long_t addr, compat_long_t data); -#endif /* __ARCH_WANT_COMPAT_SYS_PTRACE */ /* * epoll (fs/eventpoll.c) compat bits follow ... diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 1e68e4c39e2c..4c8bcd7dd8e0 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -612,7 +612,7 @@ int generic_ptrace_pokedata(struct task_struct *tsk, long addr, long data) return (copied == sizeof(data)) ? 0 : -EIO; } -#if defined CONFIG_COMPAT && defined __ARCH_WANT_COMPAT_SYS_PTRACE +#if defined CONFIG_COMPAT #include int compat_ptrace_request(struct task_struct *child, compat_long_t request, @@ -709,4 +709,4 @@ asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid, unlock_kernel(); return ret; } -#endif /* CONFIG_COMPAT && __ARCH_WANT_COMPAT_SYS_PTRACE */ +#endif /* CONFIG_COMPAT */ -- cgit v1.2.3 From eac9fbc6a90ab3440f4d98a8c52b15a724fc6f4f Mon Sep 17 00:00:00 2001 From: Richard Kennedy Date: Mon, 24 Nov 2008 13:53:24 +0000 Subject: AMD IOMMU: struct amd_iommu remove padding on 64 bit Remove 16 bytes of padding from struct amd_iommu on 64bit builds reducing its size to 120 bytes, allowing it to span one fewer cachelines. Signed-off-by: Richard Kennedy Signed-off-by: Joerg Roedel --- arch/x86/include/asm/amd_iommu_types.h | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index 1a30c0440c6b..ac302a2fa339 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -251,13 +251,6 @@ struct amd_iommu { /* Pointer to PCI device of this IOMMU */ struct pci_dev *dev; - /* - * Capability pointer. There could be more than one IOMMU per PCI - * device function if there are more than one AMD IOMMU capability - * pointers. - */ - u16 cap_ptr; - /* physical address of MMIO space */ u64 mmio_phys; /* virtual address of MMIO space */ @@ -266,6 +259,13 @@ struct amd_iommu { /* capabilities of that IOMMU read from ACPI */ u32 cap; + /* + * Capability pointer. There could be more than one IOMMU per PCI + * device function if there are more than one AMD IOMMU capability + * pointers. + */ + u16 cap_ptr; + /* pci domain of this IOMMU */ u16 pci_seg; @@ -284,19 +284,19 @@ struct amd_iommu { /* size of command buffer */ u32 cmd_buf_size; - /* event buffer virtual address */ - u8 *evt_buf; /* size of event buffer */ u32 evt_buf_size; + /* event buffer virtual address */ + u8 *evt_buf; /* MSI number for event interrupt */ u16 evt_msi_num; - /* if one, we need to send a completion wait command */ - int need_sync; - /* true if interrupts for this IOMMU are already enabled */ bool int_enabled; + /* if one, we need to send a completion wait command */ + int need_sync; + /* default dma_ops domain for that IOMMU */ struct dma_ops_domain *default_dom; }; -- cgit v1.2.3 From f91ba190648be4ff127d6aaf3993ac19d66dc2c2 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 25 Nov 2008 12:56:12 +0100 Subject: AMD IOMMU: set device table entry for aliased devices In some rare cases a request can arrive an IOMMU with its originial requestor id even it is aliased. Handle this by setting the device table entry to the same protection domain for the original and the aliased requestor id. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index e4899e0e8787..a232e5a85d48 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -858,6 +858,9 @@ static int get_device_resources(struct device *dev, print_devid(_bdf, 1); } + if (domain_for_device(_bdf) == NULL) + set_device_domain(*iommu, *domain, _bdf); + return 1; } -- cgit v1.2.3 From 09ee17eb8ea89514c13980c4010bdbbaea8630c2 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 3 Dec 2008 12:19:27 +0100 Subject: AMD IOMMU: fix possible race while accessing iommu->need_sync The access to the iommu->need_sync member needs to be protected by the iommu->lock. Otherwise this is a possible race condition. Fix it with this patch. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 33 +++++++++++++-------------------- 1 file changed, 13 insertions(+), 20 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index a232e5a85d48..5662e226b0c9 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -187,6 +187,8 @@ static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd) spin_lock_irqsave(&iommu->lock, flags); ret = __iommu_queue_command(iommu, cmd); + if (!ret) + iommu->need_sync = 1; spin_unlock_irqrestore(&iommu->lock, flags); return ret; @@ -210,10 +212,13 @@ static int iommu_completion_wait(struct amd_iommu *iommu) cmd.data[0] = CMD_COMPL_WAIT_INT_MASK; CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT); - iommu->need_sync = 0; - spin_lock_irqsave(&iommu->lock, flags); + if (!iommu->need_sync) + goto out; + + iommu->need_sync = 0; + ret = __iommu_queue_command(iommu, &cmd); if (ret) @@ -254,8 +259,6 @@ static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid) ret = iommu_queue_command(iommu, &cmd); - iommu->need_sync = 1; - return ret; } @@ -281,8 +284,6 @@ static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu, ret = iommu_queue_command(iommu, &cmd); - iommu->need_sync = 1; - return ret; } @@ -762,8 +763,6 @@ static void set_device_domain(struct amd_iommu *iommu, write_unlock_irqrestore(&amd_iommu_devtable_lock, flags); iommu_queue_inv_dev_entry(iommu, devid); - - iommu->need_sync = 1; } /***************************************************************************** @@ -1034,8 +1033,7 @@ static dma_addr_t map_single(struct device *dev, phys_addr_t paddr, if (addr == bad_dma_address) goto out; - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); out: spin_unlock_irqrestore(&domain->lock, flags); @@ -1063,8 +1061,7 @@ static void unmap_single(struct device *dev, dma_addr_t dma_addr, __unmap_single(iommu, domain->priv, dma_addr, size, dir); - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); spin_unlock_irqrestore(&domain->lock, flags); } @@ -1130,8 +1127,7 @@ static int map_sg(struct device *dev, struct scatterlist *sglist, goto unmap; } - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); out: spin_unlock_irqrestore(&domain->lock, flags); @@ -1176,8 +1172,7 @@ static void unmap_sg(struct device *dev, struct scatterlist *sglist, s->dma_address = s->dma_length = 0; } - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); spin_unlock_irqrestore(&domain->lock, flags); } @@ -1228,8 +1223,7 @@ static void *alloc_coherent(struct device *dev, size_t size, goto out; } - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); out: spin_unlock_irqrestore(&domain->lock, flags); @@ -1260,8 +1254,7 @@ static void free_coherent(struct device *dev, size_t size, __unmap_single(iommu, domain->priv, dma_addr, size, DMA_BIDIRECTIONAL); - if (unlikely(iommu->need_sync)) - iommu_completion_wait(iommu); + iommu_completion_wait(iommu); spin_unlock_irqrestore(&domain->lock, flags); -- cgit v1.2.3