From 1cc1c96c1658bfaf85d06d764bd7ac00640ae90f Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 20 Feb 2012 17:23:47 -0800 Subject: PCI: fix memleak when ACPI _CRS is not used. warning: unreferenced object 0xffff8801f6914200 (size 512): comm "swapper/0", pid 1, jiffies 4294893643 (age 2664.644s) hex dump (first 32 bytes): 00 00 c0 fe 00 00 00 00 ff ff ff ff 00 00 00 00 ................ 60 58 2f f6 03 88 ff ff 00 02 00 00 00 00 00 00 `X/............. backtrace: [] kmemleak_alloc+0x26/0x43 [] __kmalloc+0x121/0x183 [] get_current_resources+0x5a/0xc6 [] pci_acpi_scan_root+0x13c/0x21c [] acpi_pci_root_add+0x1e1/0x421 [] acpi_device_probe+0x50/0x190 [] really_probe+0x99/0x126 [] driver_probe_device+0x3b/0x56 [] __driver_attach+0x5f/0x82 [] bus_for_each_dev+0x5c/0x88 [] driver_attach+0x1e/0x20 [] bus_add_driver+0xca/0x21d [] driver_register+0x91/0xfe [] acpi_bus_register_driver+0x43/0x45 [] acpi_pci_root_init+0x20/0x28 [] do_one_initcall+0x57/0x134 The system has _CRS for root buses, but they are not used because the machine date is before the cutoff date for _CRS usage. Try to free those unused resource arrays and names. Reviewed-by: Bjorn Helgaas Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes --- arch/x86/pci/acpi.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c index a312e76063a7..c33e0970ee9f 100644 --- a/arch/x86/pci/acpi.c +++ b/arch/x86/pci/acpi.c @@ -282,9 +282,6 @@ static void add_resources(struct pci_root_info *info) int i; struct resource *res, *root, *conflict; - if (!pci_use_crs) - return; - coalesce_windows(info, IORESOURCE_MEM); coalesce_windows(info, IORESOURCE_IO); @@ -336,8 +333,13 @@ get_current_resources(struct acpi_device *device, int busnum, acpi_walk_resources(device->handle, METHOD_NAME__CRS, setup_resource, &info); - add_resources(&info); - return; + if (pci_use_crs) { + add_resources(&info); + + return; + } + + kfree(info.name); name_alloc_fail: kfree(info.res); -- cgit v1.2.3 From 8411371709610c826bf65684f886bfdfb5780ca1 Mon Sep 17 00:00:00 2001 From: Jonathan Nieder Date: Tue, 28 Feb 2012 11:51:10 -0700 Subject: x86/PCI: use host bridge _CRS info on MSI MS-7253 In the spirit of commit 29cf7a30f8a0 ("x86/PCI: use host bridge _CRS info on ASUS M2V-MX SE"), this DMI quirk turns on "pci_use_crs" by default on a board that needs it. This fixes boot failures and oopses introduced in 3e3da00c01d0 ("x86/pci: AMD one chain system to use pci read out res"). The quirk is quite targetted (to a specific board and BIOS version) for two reasons: (1) to emphasize that this method of tackling the problem one quirk at a time is a little insane (2) to give BIOS vendors an opportunity to use simpler tables and allow us to return to generic behavior (whatever that happens to be) with a later BIOS update In other words, I am not at all happy with having quirks like this. But it is even worse for the kernel not to work out of the box on these machines, so... Reference: https://bugzilla.kernel.org/show_bug.cgi?id=42619 Reported-by: Svante Signell Signed-off-by: Jonathan Nieder Signed-off-by: Bjorn Helgaas Signed-off-by: Jesse Barnes --- arch/x86/pci/acpi.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c index c33e0970ee9f..7034c081b226 100644 --- a/arch/x86/pci/acpi.c +++ b/arch/x86/pci/acpi.c @@ -60,6 +60,17 @@ static const struct dmi_system_id pci_use_crs_table[] __initconst = { DMI_MATCH(DMI_BIOS_VENDOR, "American Megatrends Inc."), }, }, + /* https://bugzilla.kernel.org/show_bug.cgi?id=42619 */ + { + .callback = set_use_crs, + .ident = "MSI MS-7253", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "MICRO-STAR INTERNATIONAL CO., LTD"), + DMI_MATCH(DMI_BOARD_NAME, "MS-7253"), + DMI_MATCH(DMI_BIOS_VENDOR, "Phoenix Technologies, LTD"), + DMI_MATCH(DMI_BIOS_VERSION, "V1.6"), + }, + }, /* Now for the blacklist.. */ -- cgit v1.2.3 From a97f4f5e524bcd09a85ef0b8821a14d35e69335f Mon Sep 17 00:00:00 2001 From: Jonathan Nieder Date: Tue, 28 Feb 2012 15:31:35 -0600 Subject: x86/PCI: do not tie MSI MS-7253 use_crs quirk to BIOS version Carlos was getting WARNING: at drivers/pci/pci.c:118 pci_ioremap_bar+0x24/0x52() when probing his sound card, and sound did not work. After adding pci=use_crs to the kernel command line, no more trouble. Ok, we can add a quirk. dmidecode output reveals that this is an MSI MS-7253, for which we already have a quirk, but the short-sighted author tied the quirk to a single BIOS version, making it not kick in on Carlos's machine with BIOS V1.2. If a later BIOS update makes it no longer necessary to look at the _CRS info it will still be harmless, so let's stop trying to guess which versions have and don't have accurate _CRS tables. Addresses https://bugtrack.alsa-project.org/alsa-bug/view.php?id=5533 Also see . Reported-by: Carlos Luna Reviewed-by: Bjorn Helgaas Signed-off-by: Jonathan Nieder Signed-off-by: Jesse Barnes --- arch/x86/pci/acpi.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c index 7034c081b226..49a5cb55429b 100644 --- a/arch/x86/pci/acpi.c +++ b/arch/x86/pci/acpi.c @@ -68,7 +68,6 @@ static const struct dmi_system_id pci_use_crs_table[] __initconst = { DMI_MATCH(DMI_BOARD_VENDOR, "MICRO-STAR INTERNATIONAL CO., LTD"), DMI_MATCH(DMI_BOARD_NAME, "MS-7253"), DMI_MATCH(DMI_BIOS_VENDOR, "Phoenix Technologies, LTD"), - DMI_MATCH(DMI_BIOS_VERSION, "V1.6"), }, }, -- cgit v1.2.3 From 6414fa6a150111750011f477899d370244da4171 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 5 Mar 2012 06:38:42 +0000 Subject: aout: move setup_arg_pages() prior to reading/mapping the binary Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/ia32/ia32_aout.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index fd843877e841..39e49091f648 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -315,6 +315,13 @@ static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs) current->mm->free_area_cache = TASK_UNMAPPED_BASE; current->mm->cached_hole_size = 0; + retval = setup_arg_pages(bprm, IA32_STACK_TOP, EXSTACK_DEFAULT); + if (retval < 0) { + /* Someone check-me: is this error path enough? */ + send_sig(SIGKILL, current, 0); + return retval; + } + install_exec_creds(bprm); current->flags &= ~PF_FORKNOEXEC; @@ -410,13 +417,6 @@ beyond_if: set_brk(current->mm->start_brk, current->mm->brk); - retval = setup_arg_pages(bprm, IA32_STACK_TOP, EXSTACK_DEFAULT); - if (retval < 0) { - /* Someone check-me: is this error path enough? */ - send_sig(SIGKILL, current, 0); - return retval; - } - current->mm->start_stack = (unsigned long)create_aout_tables((char __user *)bprm->p, bprm); /* start thread */ -- cgit v1.2.3 From 097d59106a8e4b42d07c9892fdd7790f1659c6ff Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 6 Mar 2012 18:23:36 -0800 Subject: vm: avoid using find_vma_prev() unnecessarily Several users of "find_vma_prev()" were not in fact interested in the previous vma if there was no primary vma to be found either. And in those cases, we're much better off just using the regular "find_vma()", and then "prev" can be looked up by just checking vma->vm_prev. The find_vma_prev() semantics are fairly subtle (see Mikulas' recent commit 83cd904d271b: "mm: fix find_vma_prev"), and the whole "return prev by reference" means that it generates worse code too. Thus this "let's avoid using this inconvenient and clearly too subtle interface when we don't really have to" patch. Cc: Mikulas Patocka Cc: KOSAKI Motohiro Signed-off-by: Linus Torvalds --- arch/x86/mm/hugetlbpage.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/hugetlbpage.c b/arch/x86/mm/hugetlbpage.c index f581a18c0d4d..83e7141c3982 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -333,13 +333,15 @@ try_again: * Lookup failure means no vma is above this address, * i.e. return with success: */ - if (!(vma = find_vma_prev(mm, addr, &prev_vma))) + vma = find_vma(mm, add); + if (!vma) return addr; /* * new region fits between prev_vma->vm_end and * vma->vm_start, use it: */ + prev_vma = vma->vm_prev; if (addr + len <= vma->vm_start && (!prev_vma || (addr >= prev_vma->vm_end))) { /* remember the address as a hint for next time */ -- cgit v1.2.3 From 55062d061790b43aee01ab3f9ac57b8596254f19 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 6 Mar 2012 18:48:13 -0800 Subject: x86: fix typo in recent find_vma_prev purge It turns out that test-compiling this file on x86-64 doesn't really help, because much of it is x86-32-specific. And so I hadn't noticed the slightly over-eager removal of the 'r' from 'addr' variable despite thinking I had tested it. Signed-off-by: Linus "oopsie" Torvalds --- arch/x86/mm/hugetlbpage.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/hugetlbpage.c b/arch/x86/mm/hugetlbpage.c index 83e7141c3982..8ecbb4bba4b3 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -333,7 +333,7 @@ try_again: * Lookup failure means no vma is above this address, * i.e. return with success: */ - vma = find_vma(mm, add); + vma = find_vma(mm, addr); if (!vma) return addr; -- cgit v1.2.3 From a7f4255f906f60f72e00aad2fb000939449ff32e Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 9 Mar 2012 20:55:10 +0100 Subject: x86: Derandom delay_tsc for 64 bit Commit f0fbf0abc093 ("x86: integrate delay functions") converted delay_tsc() into a random delay generator for 64 bit. The reason is that it merged the mostly identical versions of delay_32.c and delay_64.c. Though the subtle difference of the result was: static void delay_tsc(unsigned long loops) { - unsigned bclock, now; + unsigned long bclock, now; Now the function uses rdtscl() which returns the lower 32bit of the TSC. On 32bit that's not problematic as unsigned long is 32bit. On 64 bit this fails when the lower 32bit are close to wrap around when bclock is read, because the following check if ((now - bclock) >= loops) break; evaluated to true on 64bit for e.g. bclock = 0xffffffff and now = 0 because the unsigned long (now - bclock) of these values results in 0xffffffff00000001 which is definitely larger than the loops value. That explains Tvortkos observation: "Because I am seeing udelay(500) (_occasionally_) being short, and that by delaying for some duration between 0us (yep) and 491us." Make those variables explicitely u32 again, so this works for both 32 and 64 bit. Reported-by: Tvrtko Ursulin Signed-off-by: Thomas Gleixner Cc: stable@vger.kernel.org # >= 2.6.27 Signed-off-by: Linus Torvalds --- arch/x86/lib/delay.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/lib/delay.c b/arch/x86/lib/delay.c index fc45ba887d05..e395693abdb1 100644 --- a/arch/x86/lib/delay.c +++ b/arch/x86/lib/delay.c @@ -48,9 +48,9 @@ static void delay_loop(unsigned long loops) } /* TSC based delay: */ -static void delay_tsc(unsigned long loops) +static void delay_tsc(unsigned long __loops) { - unsigned long bclock, now; + u32 bclock, now, loops = __loops; int cpu; preempt_disable(); -- cgit v1.2.3 From 87e24f4b67e68d9fd8df16e0bf9c66d1ad2a2533 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 5 Mar 2012 23:59:25 +0100 Subject: perf/x86: Fix local vs remote memory events for NHM/WSM Verified using the below proglet.. before: [root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0 remote write Performance counter stats for './numa 0': 2,101,554 node-stores 2,096,931 node-store-misses 5.021546079 seconds time elapsed [root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1 local write Performance counter stats for './numa 1': 501,137 node-stores 199 node-store-misses 5.124451068 seconds time elapsed After: [root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 0 remote write Performance counter stats for './numa 0': 2,107,516 node-stores 2,097,187 node-store-misses 5.012755149 seconds time elapsed [root@westmere ~]# perf stat -e node-stores -e node-store-misses ./numa 1 local write Performance counter stats for './numa 1': 2,063,355 node-stores 165 node-store-misses 5.082091494 seconds time elapsed #define _GNU_SOURCE #include #include #include #include #include #include #include #include #include #include #define SIZE (32*1024*1024) volatile int done; void sig_done(int sig) { done = 1; } int main(int argc, char **argv) { cpu_set_t *mask, *mask2; size_t size; int i, err, t; int nrcpus = 1024; char *mem; unsigned long nodemask = 0x01; /* node 0 */ DIR *node; struct dirent *de; int read = 0; int local = 0; if (argc < 2) { printf("usage: %s [0-3]\n", argv[0]); printf(" bit0 - local/remote\n"); printf(" bit1 - read/write\n"); exit(0); } switch (atoi(argv[1])) { case 0: printf("remote write\n"); break; case 1: printf("local write\n"); local = 1; break; case 2: printf("remote read\n"); read = 1; break; case 3: printf("local read\n"); local = 1; read = 1; break; } mask = CPU_ALLOC(nrcpus); size = CPU_ALLOC_SIZE(nrcpus); CPU_ZERO_S(size, mask); node = opendir("/sys/devices/system/node/node0/"); if (!node) perror("opendir"); while ((de = readdir(node))) { int cpu; if (sscanf(de->d_name, "cpu%d", &cpu) == 1) CPU_SET_S(cpu, size, mask); } closedir(node); mask2 = CPU_ALLOC(nrcpus); CPU_ZERO_S(size, mask2); for (i = 0; i < size; i++) CPU_SET_S(i, size, mask2); CPU_XOR_S(size, mask2, mask2, mask); // invert if (!local) mask = mask2; err = sched_setaffinity(0, size, mask); if (err) perror("sched_setaffinity"); mem = mmap(0, SIZE, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0); err = mbind(mem, SIZE, MPOL_BIND, &nodemask, 8*sizeof(nodemask), MPOL_MF_MOVE); if (err) perror("mbind"); signal(SIGALRM, sig_done); alarm(5); if (!read) { while (!done) { for (i = 0; i < SIZE; i++) mem[i] = 0x01; } } else { while (!done) { for (i = 0; i < SIZE; i++) t += *(volatile char *)(mem + i); } } return 0; } Signed-off-by: Peter Zijlstra Cc: Stephane Eranian Cc: Link: http://lkml.kernel.org/n/tip-tq73sxus35xmqpojf7ootxgs@git.kernel.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 3bd37bdf1b8e..61d4f79a550e 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -385,14 +385,15 @@ static __initconst const u64 westmere_hw_cache_event_ids #define NHM_LOCAL_DRAM (1 << 14) #define NHM_NON_DRAM (1 << 15) -#define NHM_ALL_DRAM (NHM_REMOTE_DRAM|NHM_LOCAL_DRAM) +#define NHM_LOCAL (NHM_LOCAL_DRAM|NHM_REMOTE_CACHE_FWD) +#define NHM_REMOTE (NHM_REMOTE_DRAM) #define NHM_DMND_READ (NHM_DMND_DATA_RD) #define NHM_DMND_WRITE (NHM_DMND_RFO|NHM_DMND_WB) #define NHM_DMND_PREFETCH (NHM_PF_DATA_RD|NHM_PF_DATA_RFO) #define NHM_L3_HIT (NHM_UNCORE_HIT|NHM_OTHER_CORE_HIT_SNP|NHM_OTHER_CORE_HITM) -#define NHM_L3_MISS (NHM_NON_DRAM|NHM_ALL_DRAM|NHM_REMOTE_CACHE_FWD) +#define NHM_L3_MISS (NHM_NON_DRAM|NHM_LOCAL_DRAM|NHM_REMOTE_DRAM|NHM_REMOTE_CACHE_FWD) #define NHM_L3_ACCESS (NHM_L3_HIT|NHM_L3_MISS) static __initconst const u64 nehalem_hw_cache_extra_regs @@ -416,16 +417,16 @@ static __initconst const u64 nehalem_hw_cache_extra_regs }, [ C(NODE) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_READ|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_READ|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_READ|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_READ|NHM_REMOTE, }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_WRITE|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_WRITE|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_WRITE|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_WRITE|NHM_REMOTE, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_PREFETCH|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_PREFETCH|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_PREFETCH|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_PREFETCH|NHM_REMOTE, }, }, }; -- cgit v1.2.3