From 4db96cf077aa938b11fe7ac79ecc9b29ec00fbab Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Wed, 16 Sep 2009 11:50:14 +0200 Subject: HWPOISON: Add PR_MCE_KILL prctl to control early kill behaviour per process This allows processes to override their early/late kill behaviour on hardware memory errors. Typically applications which are memory error aware is better of with early kill (see the error as soon as possible), all others with late kill (only see the error when the error is really impacting execution) There's a global sysctl, but this way an application can set its specific policy. We're using two bits, one to signify that the process stated its intention and that I also made the prctl future proof by enforcing the unused arguments are 0. The state is inherited to children. Note this makes us officially run out of process flags on 32bit, but the next patch can easily add another field. Manpage patch will be supplied separately. Signed-off-by: Andi Kleen --- kernel/sys.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'kernel/sys.c') diff --git a/kernel/sys.c b/kernel/sys.c index b3f1097c76fa..41e02eff3398 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1528,6 +1528,28 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, current->timer_slack_ns = arg2; error = 0; break; + case PR_MCE_KILL: + if (arg4 | arg5) + return -EINVAL; + switch (arg2) { + case 0: + if (arg3 != 0) + return -EINVAL; + current->flags &= ~PF_MCE_PROCESS; + break; + case 1: + current->flags |= PF_MCE_PROCESS; + if (arg3 != 0) + current->flags |= PF_MCE_EARLY; + else + current->flags &= ~PF_MCE_EARLY; + break; + default: + return -EINVAL; + } + error = 0; + break; + default: error = -EINVAL; break; -- cgit v1.2.2 From cdd6c482c9ff9c55475ee7392ec8f672eddb7be6 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Mon, 21 Sep 2009 12:02:48 +0200 Subject: perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian Acked-by: Peter Zijlstra Acked-by: Paul Mackerras Reviewed-by: Arjan van de Ven Cc: Mike Galbraith Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Steven Rostedt Cc: Benjamin Herrenschmidt Cc: David Howells Cc: Kyle McMartin Cc: Martin Schwidefsky Cc: "David S. Miller" Cc: Thomas Gleixner Cc: "H. Peter Anvin" Cc: LKML-Reference: Signed-off-by: Ingo Molnar --- kernel/sys.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel/sys.c') diff --git a/kernel/sys.c b/kernel/sys.c index b3f1097c76fa..ea5c3bcac881 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -14,7 +14,7 @@ #include #include #include -#include +#include #include #include #include @@ -1511,11 +1511,11 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, case PR_SET_TSC: error = SET_TSC_CTL(arg2); break; - case PR_TASK_PERF_COUNTERS_DISABLE: - error = perf_counter_task_disable(); + case PR_TASK_PERF_EVENTS_DISABLE: + error = perf_event_task_disable(); break; - case PR_TASK_PERF_COUNTERS_ENABLE: - error = perf_counter_task_enable(); + case PR_TASK_PERF_EVENTS_ENABLE: + error = perf_event_task_enable(); break; case PR_GET_TIMERSLACK: error = current->timer_slack_ns; -- cgit v1.2.2 From 1f10206cf8e945220f7220a809d8bfc15c21f9a5 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 22 Sep 2009 16:44:10 -0700 Subject: getrusage: fill ru_maxrss value Make ->ru_maxrss value in struct rusage filled accordingly to rss hiwater mark. This struct is filled as a parameter to getrusage syscall. ->ru_maxrss value is set to KBs which is the way it is done in BSD systems. /usr/bin/time (gnu time) application converts ->ru_maxrss to KBs which seems to be incorrect behavior. Maintainer of this util was notified by me with the patch which corrects it and cc'ed. To make this happen we extend struct signal_struct by two fields. The first one is ->maxrss which we use to store rss hiwater of the task. The second one is ->cmaxrss which we use to store highest rss hiwater of all task childs. These values are used in k_getrusage() to actually fill ->ru_maxrss. k_getrusage() uses current rss hiwater value directly if mm struct exists. Note: exec() clear mm->hiwater_rss, but doesn't clear sig->maxrss. it is intetionally behavior. *BSD getrusage have exec() inheriting. test programs ======================================================== getrusage.c =========== #include #include #include #include #include #include #include #include #include #include #include #include "common.h" #define err(str) perror(str), exit(1) int main(int argc, char** argv) { int status; printf("allocate 100MB\n"); consume(100); printf("testcase1: fork inherit? \n"); printf(" expect: initial.self ~= child.self\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase2: fork inherit? (cont.) \n"); printf(" expect: initial.children ~= 100MB, but child.children = 0\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { show_rusage("child"); _exit(0); } printf("\n"); printf("testcase3: fork + malloc \n"); printf(" expect: child.self ~= initial.self + 50MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); } else { printf("allocate +50MB\n"); consume(50); show_rusage("fork child"); _exit(0); } printf("\n"); printf("testcase4: grandchild maxrss\n"); printf(" expect: post_wait.children ~= 300MB\n"); show_rusage("initial"); if (__fork()) { wait(&status); show_rusage("post_wait"); } else { system("./child -n 0 -g 300"); _exit(0); } printf("\n"); printf("testcase5: zombie\n"); printf(" expect: pre_wait ~= initial, IOW the zombie process is not accounted.\n"); printf(" post_wait ~= 400MB, IOW wait() collect child's max_rss. \n"); show_rusage("initial"); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("pre_wait"); wait(&status); show_rusage("post_wait"); } else { system("./child -n 400"); _exit(0); } printf("\n"); printf("testcase6: SIG_IGN\n"); printf(" expect: initial ~= after_zombie (child's 500MB alloc should be ignored).\n"); show_rusage("initial"); signal(SIGCHLD, SIG_IGN); if (__fork()) { sleep(1); /* children become zombie */ show_rusage("after_zombie"); } else { system("./child -n 500"); _exit(0); } printf("\n"); signal(SIGCHLD, SIG_DFL); printf("testcase7: exec (without fork) \n"); printf(" expect: initial ~= exec \n"); show_rusage("initial"); execl("./child", "child", "-v", NULL); return 0; } child.c ======= #include #include #include #include #include #include #include #include #include #include #include "common.h" int main(int argc, char** argv) { int status; int c; long consume_size = 0; long grandchild_consume_size = 0; int show = 0; while ((c = getopt(argc, argv, "n:g:v")) != -1) { switch (c) { case 'n': consume_size = atol(optarg); break; case 'v': show = 1; break; case 'g': grandchild_consume_size = atol(optarg); break; default: break; } } if (show) show_rusage("exec"); if (consume_size) { printf("child alloc %ldMB\n", consume_size); consume(consume_size); } if (grandchild_consume_size) { if (fork()) { wait(&status); } else { printf("grandchild alloc %ldMB\n", grandchild_consume_size); consume(grandchild_consume_size); exit(0); } } return 0; } common.c ======== #include #include #include #include #include #include #include #include #include #include #include #include "common.h" #define err(str) perror(str), exit(1) void show_rusage(char *prefix) { int err, err2; struct rusage rusage_self; struct rusage rusage_children; printf("%s: ", prefix); err = getrusage(RUSAGE_SELF, &rusage_self); if (!err) printf("self %ld ", rusage_self.ru_maxrss); err2 = getrusage(RUSAGE_CHILDREN, &rusage_children); if (!err2) printf("children %ld ", rusage_children.ru_maxrss); printf("\n"); } /* Some buggy OS need this worthless CPU waste. */ void make_pagefault(void) { void *addr; int size = getpagesize(); int i; for (i=0; i<1000; i++) { addr = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0); if (addr == MAP_FAILED) err("make_pagefault"); memset(addr, 0, size); munmap(addr, size); } } void consume(int mega) { size_t sz = mega * 1024 * 1024; void *ptr; ptr = malloc(sz); memset(ptr, 0, sz); make_pagefault(); } pid_t __fork(void) { pid_t pid; pid = fork(); make_pagefault(); return pid; } common.h ======== void show_rusage(char *prefix); void make_pagefault(void); void consume(int mega); pid_t __fork(void); FreeBSD result (expected result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 103492 children 0 fork child: self 103540 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 103540 children 103540 child: self 103564 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 103564 children 103564 allocate +50MB fork child: self 154860 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 103564 children 154860 grandchild alloc 300MB post_wait: self 103564 children 308720 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 103564 children 308720 child alloc 400MB pre_wait: self 103564 children 308720 post_wait: self 103564 children 411312 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 103564 children 411312 child alloc 500MB after_zombie: self 103624 children 411312 testcase7: exec (without fork) expect: initial ~= exec initial: self 103624 children 411312 exec: self 103624 children 411312 Linux result (actual test result) ======================================================== allocate 100MB testcase1: fork inherit? expect: initial.self ~= child.self initial: self 102848 children 0 fork child: self 102572 children 0 testcase2: fork inherit? (cont.) expect: initial.children ~= 100MB, but child.children = 0 initial: self 102876 children 102644 child: self 102572 children 0 testcase3: fork + malloc expect: child.self ~= initial.self + 50MB initial: self 102876 children 102644 allocate +50MB fork child: self 153804 children 0 testcase4: grandchild maxrss expect: post_wait.children ~= 300MB initial: self 102876 children 153864 grandchild alloc 300MB post_wait: self 102876 children 307536 testcase5: zombie expect: pre_wait ~= initial, IOW the zombie process is not accounted. post_wait ~= 400MB, IOW wait() collect child's max_rss. initial: self 102876 children 307536 child alloc 400MB pre_wait: self 102876 children 307536 post_wait: self 102876 children 410076 testcase6: SIG_IGN expect: initial ~= after_zombie (child's 500MB alloc should be ignored). initial: self 102876 children 410076 child alloc 500MB after_zombie: self 102880 children 410076 testcase7: exec (without fork) expect: initial ~= exec initial: self 102880 children 410076 exec: self 102880 children 410076 Signed-off-by: Jiri Pirko Signed-off-by: KOSAKI Motohiro Cc: Oleg Nesterov Cc: Hugh Dickins Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sys.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'kernel/sys.c') diff --git a/kernel/sys.c b/kernel/sys.c index ea5c3bcac881..ebcb15611728 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1338,6 +1338,7 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) unsigned long flags; cputime_t utime, stime; struct task_cputime cputime; + unsigned long maxrss = 0; memset((char *) r, 0, sizeof *r); utime = stime = cputime_zero; @@ -1346,6 +1347,7 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) utime = task_utime(current); stime = task_stime(current); accumulate_thread_rusage(p, r); + maxrss = p->signal->maxrss; goto out; } @@ -1363,6 +1365,7 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) r->ru_majflt = p->signal->cmaj_flt; r->ru_inblock = p->signal->cinblock; r->ru_oublock = p->signal->coublock; + maxrss = p->signal->cmaxrss; if (who == RUSAGE_CHILDREN) break; @@ -1377,6 +1380,8 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) r->ru_majflt += p->signal->maj_flt; r->ru_inblock += p->signal->inblock; r->ru_oublock += p->signal->oublock; + if (maxrss < p->signal->maxrss) + maxrss = p->signal->maxrss; t = p; do { accumulate_thread_rusage(t, r); @@ -1392,6 +1397,15 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) out: cputime_to_timeval(utime, &r->ru_utime); cputime_to_timeval(stime, &r->ru_stime); + + if (who != RUSAGE_CHILDREN) { + struct mm_struct *mm = get_task_mm(p); + if (mm) { + setmax_mm_hiwater_rss(&maxrss, mm); + mmput(mm); + } + } + r->ru_maxrss = maxrss * (PAGE_SIZE / 1024); /* convert pages to KBs */ } int getrusage(struct task_struct *p, int who, struct rusage __user *ru) -- cgit v1.2.2