From 407984f1af259b31957c7c05075a454a751bb801 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:27 +0200 Subject: [PATCH] x86: Add abilty to enable/disable nmi watchdog with sysctl Adds a new /proc/sys/kernel/nmi call that will enable/disable the nmi watchdog. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- include/linux/sysctl.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index 736ed917a4f8..ecb79ba52ae1 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -150,6 +150,7 @@ enum KERN_IA64_UNALIGNED=72, /* int: ia64 unaligned userland trap enable */ KERN_COMPAT_LOG=73, /* int: print compat layer messages */ KERN_MAX_LOCK_DEPTH=74, + KERN_NMI_WATCHDOG=75, /* int: enable/disable nmi watchdog */ }; -- cgit v1.2.3 From 8da5adda91df3d2fcc5300e68da491694c9af019 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:27 +0200 Subject: [PATCH] x86: Allow users to force a panic on NMI To quote Alan Cox: The default Linux behaviour on an NMI of either memory or unknown is to continue operation. For many environments such as scientific computing it is preferable that the box is taken out and the error dealt with than an uncorrected parity/ECC error get propogated. A small number of systems do generate NMI's for bizarre random reasons such as power management so the default is unchanged. In other respects the new proc/sys entry works like the existing panic controls already in that directory. This is separate to the edac support - EDAC allows supported chipsets to handle ECC errors well, this change allows unsupported cases to at least panic rather than cause problems further down the line. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- include/linux/kernel.h | 1 + include/linux/sysctl.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 2b2ae4fdce8b..1ff9609300b4 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -186,6 +186,7 @@ extern void bust_spinlocks(int yes); extern int oops_in_progress; /* If set, an oops, panic(), BUG() or die() is in progress */ extern int panic_timeout; extern int panic_on_oops; +extern int panic_on_unrecovered_nmi; extern int tainted; extern const char *print_tainted(void); extern void add_taint(unsigned); diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index ecb79ba52ae1..432778446ad2 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -151,6 +151,7 @@ enum KERN_COMPAT_LOG=73, /* int: print compat layer messages */ KERN_MAX_LOCK_DEPTH=74, KERN_NMI_WATCHDOG=75, /* int: enable/disable nmi watchdog */ + KERN_PANIC_ON_NMI=76, /* int: whether we will panic on an unrecovered */ }; -- cgit v1.2.3 From c08c820508233b424deab3302bc404bbecc6493a Mon Sep 17 00:00:00 2001 From: Vojtech Pavlik Date: Tue, 26 Sep 2006 10:52:28 +0200 Subject: [PATCH] Add the vgetcpu vsyscall This patch adds a vgetcpu vsyscall, which depending on the CPU RDTSCP capability uses either the RDTSCP or CPUID to obtain a CPU and node numbers and pass them to the program. AK: Lots of changes over Vojtech's original code: Better prototype for vgetcpu() It's better to pass the cpu / node numbers as separate arguments to avoid mistakes when going from SMP to NUMA. Also add a fast time stamp based cache using a user supplied argument to speed things more up. Use fast method from Chuck Ebbert to retrieve node/cpu from GDT limit instead of CPUID Made sure RDTSCP init is always executed after node is known. Drop printk Signed-off-by: Vojtech Pavlik Signed-off-by: Andi Kleen --- include/linux/getcpu.h | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) create mode 100644 include/linux/getcpu.h (limited to 'include/linux') diff --git a/include/linux/getcpu.h b/include/linux/getcpu.h new file mode 100644 index 000000000000..031ed3780e45 --- /dev/null +++ b/include/linux/getcpu.h @@ -0,0 +1,16 @@ +#ifndef _LINUX_GETCPU_H +#define _LINUX_GETCPU_H 1 + +/* Cache for getcpu() to speed it up. Results might be upto a jiffie + out of date, but will be faster. + User programs should not refer to the contents of this structure. + It is only a cache for vgetcpu(). It might change in future kernels. + The user program must store this information per thread (__thread) + If you want 100% accurate information pass NULL instead. */ +struct getcpu_cache { + unsigned long t0; + unsigned long t1; + unsigned long res[4]; +}; + +#endif -- cgit v1.2.3 From 3cfc348bf90ffaa777c188652aa297f04eb94de8 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:28 +0200 Subject: [PATCH] x86: Add portable getcpu call For NUMA optimization and some other algorithms it is useful to have a fast to get the current CPU and node numbers in user space. x86-64 added a fast way to do this in a vsyscall. This adds a generic syscall for other architectures to make it a generic portable facility. I expect some of them will also implement it as a faster vsyscall. The cache is an optimization for the x86-64 vsyscall optimization. Since what the syscall returns is an approximation anyways and user space often wants very fast results it can be cached for some time. The norma methods to get this information in user space are relatively slow The vsyscall is in a better position to manage the cache because it has direct access to a fast time stamp (jiffies). For the generic syscall optimization it doesn't help much, but enforce a valid argument to keep programs portable I only added an i386 syscall entry for now. Other architectures can follow as needed. AK: Also added some cleanups from Andrew Morton Signed-off-by: Andi Kleen --- include/linux/syscalls.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 008f04c56737..3f0f716225ec 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -53,6 +53,7 @@ struct mq_attr; struct compat_stat; struct compat_timeval; struct robust_list_head; +struct getcpu_cache; #include #include @@ -596,5 +597,6 @@ asmlinkage long sys_get_robust_list(int pid, size_t __user *len_ptr); asmlinkage long sys_set_robust_list(struct robust_list_head __user *head, size_t len); +asmlinkage long sys_getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *cache); #endif -- cgit v1.2.3 From 5a1b3999d6cb7ab87f1f3b1700bc91839fd6fa29 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:34 +0200 Subject: [PATCH] x86: Some preparationary cleanup for stack trace - Remove unused all_contexts parameter No caller used it - Move skip argument into the structure (needed for followon patches) Cc: mingo@elte.hu Signed-off-by: Andi Kleen --- include/linux/stacktrace.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/stacktrace.h b/include/linux/stacktrace.h index 9cc81e572224..50e2b01e517c 100644 --- a/include/linux/stacktrace.h +++ b/include/linux/stacktrace.h @@ -5,15 +5,16 @@ struct stack_trace { unsigned int nr_entries, max_entries; unsigned long *entries; + int skip; /* input argument: How many entries to skip */ + int all_contexts; /* input argument: if true do than one stack */ }; extern void save_stack_trace(struct stack_trace *trace, - struct task_struct *task, int all_contexts, - unsigned int skip); + struct task_struct *task); extern void print_stack_trace(struct stack_trace *trace, int spaces); #else -# define save_stack_trace(trace, task, all, skip) do { } while (0) +# define save_stack_trace(trace, task) do { } while (0) # define print_stack_trace(trace) do { } while (0) #endif -- cgit v1.2.3 From d28c4393a7bf558538e9def269c1caeab6ec056f Mon Sep 17 00:00:00 2001 From: "Prasanna S.P" Date: Tue, 26 Sep 2006 10:52:34 +0200 Subject: [PATCH] x86: error_code is not safe for kprobes This patch moves the entry.S:error_entry to .kprobes.text section, since code marked unsafe for kprobes jumps directly to entry.S::error_entry, that must be marked unsafe as well. This patch also moves all the ".previous.text" asm directives to ".previous" for kprobes section. AK: Following a similar i386 patch from Chuck Ebbert AK: Also merged Jeremy's fix in. +From: Jeremy Fitzhardinge KPROBE_ENTRY does a .section .kprobes.text, and expects its users to do a .previous at the end of the function. Unfortunately, if any code within the function switches sections, for example .fixup, then the .previous ends up putting all subsequent code into .fixup. Worse, any subsequent .fixup code gets intermingled with the code its supposed to be fixing (which is also in .fixup). It's surprising this didn't cause more havok. The fix is to use .pushsection/.popsection, so this stuff nests properly. A further cleanup would be to get rid of all .section/.previous pairs, since they're inherently fragile. +From: Chuck Ebbert <76306.1226@compuserve.com> Because code marked unsafe for kprobes jumps directly to entry.S::error_code, that must be marked unsafe as well. The easiest way to do that is to move the page fault entry point to just before error_code and let it inherit the same section. Also moved all the ".previous" asm directives for kprobes sections to column 1 and removed ".text" from them. Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com> Signed-off-by: Andi Kleen --- include/linux/linkage.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/linkage.h b/include/linux/linkage.h index 932021f872d5..6c9873f88287 100644 --- a/include/linux/linkage.h +++ b/include/linux/linkage.h @@ -35,9 +35,13 @@ #endif #define KPROBE_ENTRY(name) \ - .section .kprobes.text, "ax"; \ + .pushsection .kprobes.text, "ax"; \ ENTRY(name) +#define KPROBE_END(name) \ + END(name); \ + .popsection + #ifndef END #define END(name) \ .size name, .-name -- cgit v1.2.3 From e07e23e1fd3000289fc7ccc6c71879070d3b19e0 Mon Sep 17 00:00:00 2001 From: Arjan van de Ven Date: Tue, 26 Sep 2006 10:52:36 +0200 Subject: [PATCH] non lazy "sleazy" fpu implementation Right now the kernel on x86-64 has a 100% lazy fpu behavior: after *every* context switch a trap is taken for the first FPU use to restore the FPU context lazily. This is of course great for applications that have very sporadic or no FPU use (since then you avoid doing the expensive save/restore all the time). However for very frequent FPU users... you take an extra trap every context switch. The patch below adds a simple heuristic to this code: After 5 consecutive context switches of FPU use, the lazy behavior is disabled and the context gets restored every context switch. If the app indeed uses the FPU, the trap is avoided. (the chance of the 6th time slice using FPU after the previous 5 having done so are quite high obviously). After 256 switches, this is reset and lazy behavior is returned (until there are 5 consecutive ones again). The reason for this is to give apps that do longer bursts of FPU use still the lazy behavior back after some time. [akpm@osdl.org: place new task_struct field next to jit_keyring to save space] Signed-off-by: Arjan van de Ven Signed-off-by: Andi Kleen Cc: Andi Kleen Signed-off-by: Andrew Morton --- include/linux/sched.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 34ed0d99b1bd..807556c5bcd2 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -865,6 +865,15 @@ struct task_struct { struct key *thread_keyring; /* keyring private to this thread */ unsigned char jit_keyring; /* default keyring to attach requested keys to */ #endif + /* + * fpu_counter contains the number of consecutive context switches + * that the FPU is used. If this is over a threshold, the lazy fpu + * saving becomes unlazy to save the trap. This is an unsigned char + * so that after 256 times the counter wraps and the behavior turns + * lazy again; this to deal with bursty apps that only use FPU for + * a short time + */ + unsigned char fpu_counter; int oomkilladj; /* OOM kill score adjustment (bit shift). */ char comm[TASK_COMM_LEN]; /* executable name excluding path - access with [gs]et_task_comm (which lock -- cgit v1.2.3 From 1bb4996bcebca1cde49d964b4e012699ce180e61 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:37 +0200 Subject: [PATCH] Move compiler check for modules to ia64 only Apparently IA64 needs it, but i386/x86-64 don't anymore since gcc 2.95 support was dropped. Nobody else on linux-arch requested keeping it generically Cc: tony.luck@intel.com Cc: kaos@sgi.com Signed-off-by: Andi Kleen --- include/linux/vermagic.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/vermagic.h b/include/linux/vermagic.h index 46919f9f5eb3..4d0909e53595 100644 --- a/include/linux/vermagic.h +++ b/include/linux/vermagic.h @@ -24,5 +24,5 @@ #define VERMAGIC_STRING \ UTS_RELEASE " " \ MODULE_VERMAGIC_SMP MODULE_VERMAGIC_PREEMPT \ - MODULE_VERMAGIC_MODULE_UNLOAD MODULE_ARCH_VERMAGIC \ - "gcc-" __stringify(__GNUC__) "." __stringify(__GNUC_MINOR__) + MODULE_VERMAGIC_MODULE_UNLOAD MODULE_ARCH_VERMAGIC + -- cgit v1.2.3 From 575400d1b483fbe9e03c68758059bfaf4e4768d1 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 26 Sep 2006 10:52:38 +0200 Subject: [PATCH] i386: Fix the EDD code misparsing the command line The EDD code would scan the command line as a fixed array, without taking account of either whitespace, null-termination, the old command-line protocol, late overrides early, or the fact that the command line may not be reachable from INITSEG. This should fix those problems, and enable us to use a longer command line. Signed-off-by: H. Peter Anvin Signed-off-by: Andi Kleen --- include/linux/edd.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/edd.h b/include/linux/edd.h index 162512b886f7..b2b3e68aa512 100644 --- a/include/linux/edd.h +++ b/include/linux/edd.h @@ -52,6 +52,7 @@ #define EDD_CL_EQUALS 0x3d646465 /* "edd=" */ #define EDD_CL_OFF 0x666f /* "of" for off */ #define EDD_CL_SKIP 0x6b73 /* "sk" for skipmbr */ +#define EDD_CL_ON 0x6e6f /* "on" for on */ #ifndef __ASSEMBLY__ -- cgit v1.2.3 From 0a4254058037eb172758961d0a5b94f4320a1425 Mon Sep 17 00:00:00 2001 From: Arjan van de Ven Date: Tue, 26 Sep 2006 10:52:38 +0200 Subject: [PATCH] Add the canary field to the PDA area and the task struct This patch adds the per thread cookie field to the task struct and the PDA. Also it makes sure that the PDA value gets the new cookie value at context switch, and that a new task gets a new cookie at task creation time. Signed-off-by: Arjan van Ven Signed-off-by: Ingo Molnar Signed-off-by: Andi Kleen CC: Andi Kleen --- include/linux/sched.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 807556c5bcd2..9d4aa7f95bc8 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -819,6 +819,11 @@ struct task_struct { unsigned did_exec:1; pid_t pid; pid_t tgid; + +#ifdef CONFIG_CC_STACKPROTECTOR + /* Canary value for the -fstack-protector gcc feature */ + unsigned long stack_canary; +#endif /* * pointers to (original) parent process, youngest child, younger sibling, * older sibling, respectively. (p->father can be replaced with -- cgit v1.2.3 From 3b171672831b9633c2ed8fa94805255cd4d5af19 Mon Sep 17 00:00:00 2001 From: Dmitriy Zavin Date: Tue, 26 Sep 2006 10:52:42 +0200 Subject: [PATCH] Add 64bit jiffies compares (for use with get_jiffies_64) The current time_before/time_after macros will fail typechecks when passed u64 values (as returned by get_jiffies_64()). On 64bit systems, this will just result in a warning about mismatching types without explicit casts, but since unsigned long and u64 (unsigned long long) are of same size, it will still work. On 32bit systems, a long is 32bits, so the value from get_jiffies_64() will be truncated by the cast and thus lose all the precision gained by 64bit jiffies. Signed-off-by: Dmitriy Zavin Signed-off-by: Andi Kleen --- include/linux/jiffies.h | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'include/linux') diff --git a/include/linux/jiffies.h b/include/linux/jiffies.h index 329ebcffa106..c8d5f207c3d4 100644 --- a/include/linux/jiffies.h +++ b/include/linux/jiffies.h @@ -115,6 +115,21 @@ static inline u64 get_jiffies_64(void) ((long)(a) - (long)(b) >= 0)) #define time_before_eq(a,b) time_after_eq(b,a) +/* Same as above, but does so with platform independent 64bit types. + * These must be used when utilizing jiffies_64 (i.e. return value of + * get_jiffies_64() */ +#define time_after64(a,b) \ + (typecheck(__u64, a) && \ + typecheck(__u64, b) && \ + ((__s64)(b) - (__s64)(a) < 0)) +#define time_before64(a,b) time_after64(b,a) + +#define time_after_eq64(a,b) \ + (typecheck(__u64, a) && \ + typecheck(__u64, b) && \ + ((__s64)(a) - (__s64)(b) >= 0)) +#define time_before_eq64(a,b) time_after_eq64(b,a) + /* * Have the 32 bit jiffies value wrap 5 minutes after boot * so jiffies wrap bugs show up earlier. -- cgit v1.2.3