From 7b1dedca42ac0d0d0be01e39d8461bb53a2389b3 Mon Sep 17 00:00:00 2001 From: Thomas Bogendoerfer Date: Sat, 29 Nov 2008 13:46:27 +0100 Subject: x86: fix dma_mapping_error for 32bit x86 Devices like b44 ethernet can't dma from addresses above 1GB. The driver handles this cases by falling back to GFP_DMA allocation. But for detecting the problem it needs to get an indication from dma_mapping_error. The bug is triggered by using a VMSPLIT option of 2G/2G. Signed-off-by: Thomas Bogendoerfer Acked-by: FUJITA Tomonori Signed-off-by: Ingo Molnar --- arch/x86/include/asm/dma-mapping.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/arch/x86/include/asm/dma-mapping.h b/arch/x86/include/asm/dma-mapping.h index 7f225a4b2a26..097794ff6b79 100644 --- a/arch/x86/include/asm/dma-mapping.h +++ b/arch/x86/include/asm/dma-mapping.h @@ -71,15 +71,13 @@ static inline struct dma_mapping_ops *get_dma_ops(struct device *dev) /* Make sure we keep the same behaviour */ static inline int dma_mapping_error(struct device *dev, dma_addr_t dma_addr) { -#ifdef CONFIG_X86_32 - return 0; -#else +#ifdef CONFIG_X86_64 struct dma_mapping_ops *ops = get_dma_ops(dev); if (ops->mapping_error) return ops->mapping_error(dev, dma_addr); - return (dma_addr == bad_dma_address); #endif + return (dma_addr == bad_dma_address); } #define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f) -- cgit v1.2.3 From 70d7d357578245f1993fd2d3ccd26088bcd38941 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 2 Dec 2008 20:16:03 +0100 Subject: x86: fix broken flushing in GART nofullflush path Impact: remove stale IOTLB entries In the non-default nofullflush case the GART is only flushed when next_bit wraps around. But it can happen that an unmap operation unmaps memory which is behind the current next_bit location. If these addresses are reused it may result in stale GART IO/TLB entries. Fix this by setting the GART next_bit always behind an unmapped location. Signed-off-by: Joerg Roedel Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-gart_64.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index a42b02b4df68..ba7ad83e20a8 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -123,6 +123,8 @@ static void free_iommu(unsigned long offset, int size) spin_lock_irqsave(&iommu_bitmap_lock, flags); iommu_area_free(iommu_gart_bitmap, offset, size); + if (offset >= next_bit) + next_bit = offset + size; spin_unlock_irqrestore(&iommu_bitmap_lock, flags); } -- 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(-) 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(+) 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(-) 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 From 9ea84ad77d635bdb76c9a08f44f21a9af98359ee Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 2 Dec 2008 07:21:21 +0100 Subject: oprofile: fix CPU unplug panic in ppro_stop() If oprofile statically compiled in kernel, a cpu unplug triggers a panic in ppro_stop(), because a NULL pointer is dereferenced. Signed-off-by: Eric Dumazet Signed-off-by: Robert Richter --- arch/x86/oprofile/op_model_ppro.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 716d26f0e5d4..e9f80c744cf3 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -156,6 +156,8 @@ static void ppro_start(struct op_msrs const * const msrs) unsigned int low, high; int i; + if (!reset_value) + return; for (i = 0; i < num_counters; ++i) { if (reset_value[i]) { CTRL_READ(low, high, msrs, i); @@ -171,6 +173,8 @@ static void ppro_stop(struct op_msrs const * const msrs) unsigned int low, high; int i; + if (!reset_value) + return; for (i = 0; i < num_counters; ++i) { if (!reset_value[i]) continue; -- cgit v1.2.3 From 3d337c653c94be50f11a45fb14a2afa8a8a1a618 Mon Sep 17 00:00:00 2001 From: William Cohen Date: Sun, 30 Nov 2008 15:39:10 -0500 Subject: x86/oprofile: fix Intel cpu family 6 detection Alan Jenkins wrote: > This is on an EeePC 701, /proc/cpuinfo as attached. > > Is this expected? Will the next release work? > > Thanks, Alan > > # opcontrol --setup --no-vmlinux > cpu_type 'unset' is not valid > you should upgrade oprofile or force the use of timer mode > > # opcontrol -v > opcontrol: oprofile 0.9.4 compiled on Nov 29 2008 22:44:10 > > # cat /dev/oprofile/cpu_type > i386/p6 > # uname -r > 2.6.28-rc6eeepc Hi Alan, Looking at the kernel driver code for oprofile it can return the "i386/p6" for the cpu_type. However, looking at the user-space oprofile code there isn't the matching entry in libop/op_cpu_type.c or the events/unit_mask files in events/i386 directory. The Intel AP-485 says this is a "Intel Pentium M processor model D". Seems like the oprofile kernel driver should be identifying the processor as "i386/p6_mobile" The driver identification code doesn't look quite right in nmi_init.c http://git.kernel.org/?p=linux/kernel/git/sfr/linux-next.git;a=blob;f=arch/x86/oprofile/nmi_int.c;h=022cd41ea9b4106e5884277096e80e9088a7c7a9;hb=HEAD has: 409 case 10 ... 13: 410 *cpu_type = "i386/p6"; 411 break; Referring to the Intel AP-485: case 10 and 11 should produce "i386/piii" case 13 should produce "i386/p6_mobile" I didn't see anything for case 12. Something like the attached patch. I don't have a celeron machine to verify that changes in this area of the kernel fix thing. -Will Signed-off-by: William Cohen Tested-by: Alan Jenkins Acked-by: Andi Kleen Signed-off-by: Robert Richter --- arch/x86/oprofile/nmi_int.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index 022cd41ea9b4..202864ad49a7 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -401,14 +401,13 @@ static int __init ppro_init(char **cpu_type) *cpu_type = "i386/pii"; break; case 6 ... 8: + case 10 ... 11: *cpu_type = "i386/piii"; break; case 9: + case 13: *cpu_type = "i386/p6_mobile"; break; - case 10 ... 13: - *cpu_type = "i386/p6"; - break; case 14: *cpu_type = "i386/core"; break; -- cgit v1.2.3 From 9adc13867ec5fe0cd35434f92954d90e42381f0b Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 4 Dec 2008 13:33:35 +0100 Subject: x86: fix early panic with boot option "nosmp" Impact: fix boot crash with numcpus=0 on certain systems Fix early exception in __get_smp_config with nosmp. Bail out early when there is no MP table. Reported-by: Wu Fengguang Signed-off-by: Andi Kleen Tested-by: Wu Fengguang Signed-off-by: Ingo Molnar --- arch/x86/kernel/mpparse.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index f98f4e1dba09..0f4c1fd5a1f4 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -604,6 +604,9 @@ static void __init __get_smp_config(unsigned int early) printk(KERN_INFO "Using ACPI for processor (LAPIC) " "configuration information\n"); + if (!mpf) + return; + printk(KERN_INFO "Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification); #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_32) -- cgit v1.2.3