mm: introduce process_mrelease system call
Subsystems (18)
PATCH v4 1/2 mm: introduce process_mrelease system call
In modern systems it's not unusual to have a system component monitoring memory conditions of the system and tasked with keeping system memory pressure under control. One way to accomplish that is to kill non-essential processes to free up memory for more important ones. Examples of this are Facebook's OOM killer daemon called oomd and Android's low memory killer daemon called lmkd.
For such system component it's important to be able to free memory quickly and efficiently. Unfortunately the time process takes to free up its memory after receiving a SIGKILL might vary based on the state of the process (uninterruptible sleep), size and OPP level of the core the process is running. A mechanism to free resources of the target process in a more predictable way would improve system's ability to control its memory pressure.
Introduce process_mrelease system call that releases memory of a dying process from the context of the caller. This way the memory is freed in a more controllable way with CPU affinity and priority of the caller. The workload of freeing the memory will also be charged to the caller. The operation is allowed only on a dying process.
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extending pidfd_send_signal to allow memory reaping using oom_reaper thread;
- https://lore.kernel.org/patchwork/patch/1338196 extending pidfd_send_signal to reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to add MADV_DONTNEED support for process_madvise implementing synchronous memory reaping.
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
The reasoning was that the new variant of process_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest of process_madvise
From the userspace point of view it was awkward and inconvenient to provide memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.
The API is as follows,
int process_mrelease(int pidfd, unsigned int flags);
DESCRIPTION
The process_mrelease() system call is used to free the memory of a process which was sent a SIGKILL signal.
The pidfd selects the process referred to by the PID file descriptor.
(See pidofd_open(2) for further information)
The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success, process_mrelease() returns 0. On error, -1 is returned and errno is set to indicate the error.
ERRORS
EBADF pidfd is not a valid PID file descriptor.
EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
EINVAL flags is not 0.
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).
ESRCH The target process does not exist (i.e., it has terminated
and been waited on).
---
changes in v4:
- Replaced mmap_read_lock() with mmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentation
mm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)
mm/oom_kill.c +58 -0
@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h>
@@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } + +SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ; +struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 ) +return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID ); +if (!task) { + ret = -ESRCH ; + goto put_pid; + } + +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task); +if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) { + ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ; +mmap_read_unlock (mm); + +put_mm: +mmput (mm); +put_task: +put_task_struct (task); +put_pid: +put_pid (pid); +return ret; +#else +return -ENOSYS ; +#endif /* CONFIG_MMU */ +}
Signature · 1 line
2.32.0.554.ge1b32706d8-goog
PATCH v4 2/2 mm: wire up syscall process_mrelease
Split off from prev patch in the series that implements the syscall.
--- arch/alpha/kernel/syscalls/syscall.tbl | 2 ++ arch/arm/tools/syscall.tbl | 2 ++ arch/arm64/include/asm/unistd.h | 2 +- arch/arm64/include/asm/unistd32.h | 2 ++ arch/ia64/kernel/syscalls/syscall.tbl | 2 ++ arch/m68k/kernel/syscalls/syscall.tbl | 2 ++ arch/microblaze/kernel/syscalls/syscall.tbl | 2 ++ arch/mips/kernel/syscalls/syscall_n32.tbl | 2 ++ arch/mips/kernel/syscalls/syscall_n64.tbl | 2 ++ arch/mips/kernel/syscalls/syscall_o32.tbl | 2 ++ arch/parisc/kernel/syscalls/syscall.tbl | 2 ++ arch/powerpc/kernel/syscalls/syscall.tbl | 2 ++ arch/s390/kernel/syscalls/syscall.tbl | 2 ++ arch/sh/kernel/syscalls/syscall.tbl | 2 ++ arch/sparc/kernel/syscalls/syscall.tbl | 2 ++ arch/x86/entry/syscalls/syscall_32.tbl | 1 + arch/x86/entry/syscalls/syscall_64.tbl | 1 + arch/xtensa/kernel/syscalls/syscall.tbl | 2 ++ include/linux/syscalls.h | 1 + include/uapi/asm-generic/unistd.h | 4 +++- kernel/sys_ni.c | 1 + 21 files changed, 38 insertions(+), 2 deletions(-)
arch/alpha/kernel/syscalls/syscall.tbl +2 -0
@@ -486,3 +486,5 @@ 554 common landlock_create_ruleset sys_landlock_create_ruleset 555 common landlock_add_rule sys_landlock_add_rule 556 common landlock_restrict_self sys_landlock_restrict_self +# 557 reserved for memfd_secret +558 common process_mrelease sys_process_mrelease
arch/arm/tools/syscall.tbl +2 -0
@@ -460,3 +460,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/arm64/include/asm/unistd.h +1 -1
@@ -38,7 +38,7 @@#define __ARM_NR_compat_set_tls (__ARM_NR_COMPAT_BASE + 5)#define __ARM_NR_COMPAT_END (__ARM_NR_COMPAT_BASE + 0x800) -#define __NR_compat_syscalls 447 +#define __NR_compat_syscalls 449#endif #define __ARCH_WANT_SYS_CLONE
arch/arm64/include/asm/unistd32.h +2 -0
@@ -901,6 +901,8 @@ __SYSCALL(__NR_landlock_create_ruleset, sys_landlock_create_ruleset) __SYSCALL(__NR_landlock_add_rule, sys_landlock_add_rule)#define __NR_landlock_restrict_self 446 __SYSCALL(__NR_landlock_restrict_self, sys_landlock_restrict_self) +#define __NR_process_mrelease 448 +__SYSCALL (__NR_process_mrelease, sys_process_mrelease)/* * Please add new compat syscalls above this comment and update
arch/ia64/kernel/syscalls/syscall.tbl +2 -0
@@ -367,3 +367,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/m68k/kernel/syscalls/syscall.tbl +2 -0
@@ -446,3 +446,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/microblaze/kernel/syscalls/syscall.tbl +2 -0
@@ -452,3 +452,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/mips/kernel/syscalls/syscall_n32.tbl +2 -0
@@ -385,3 +385,5 @@ 444 n32 landlock_create_ruleset sys_landlock_create_ruleset 445 n32 landlock_add_rule sys_landlock_add_rule 446 n32 landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 n32 process_mrelease sys_process_mrelease
arch/mips/kernel/syscalls/syscall_n64.tbl +2 -0
@@ -361,3 +361,5 @@ 444 n64 landlock_create_ruleset sys_landlock_create_ruleset 445 n64 landlock_add_rule sys_landlock_add_rule 446 n64 landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 n64 process_mrelease sys_process_mrelease
arch/mips/kernel/syscalls/syscall_o32.tbl +2 -0
@@ -434,3 +434,5 @@ 444 o32 landlock_create_ruleset sys_landlock_create_ruleset 445 o32 landlock_add_rule sys_landlock_add_rule 446 o32 landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 o32 process_mrelease sys_process_mrelease
arch/parisc/kernel/syscalls/syscall.tbl +2 -0
@@ -444,3 +444,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/powerpc/kernel/syscalls/syscall.tbl +2 -0
@@ -526,3 +526,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/s390/kernel/syscalls/syscall.tbl +2 -0
@@ -449,3 +449,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease sys_process_mrelease
arch/sh/kernel/syscalls/syscall.tbl +2 -0
@@ -449,3 +449,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/sparc/kernel/syscalls/syscall.tbl +2 -0
@@ -492,3 +492,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
arch/x86/entry/syscalls/syscall_32.tbl +1 -0
@@ -452,3 +452,4 @@ 445 i386 landlock_add_rule sys_landlock_add_rule 446 i386 landlock_restrict_self sys_landlock_restrict_self 447 i386 memfd_secret sys_memfd_secret +448 i386 process_mrelease sys_process_mrelease
arch/x86/entry/syscalls/syscall_64.tbl +1 -0
@@ -369,6 +369,7 @@ 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self 447 common memfd_secret sys_memfd_secret +448 common process_mrelease sys_process_mrelease # # Due to a historical design error, certain syscalls are numbered differently
arch/xtensa/kernel/syscalls/syscall.tbl +2 -0
@@ -417,3 +417,5 @@ 444 common landlock_create_ruleset sys_landlock_create_ruleset 445 common landlock_add_rule sys_landlock_add_rule 446 common landlock_restrict_self sys_landlock_restrict_self +# 447 reserved for memfd_secret +448 common process_mrelease sys_process_mrelease
include/linux/syscalls.h +1 -0
@@ -915,6 +915,7 @@ asmlinkage long sys_mincore(unsigned long start, size_t len,asmlinkage long sys_madvise (unsigned long start,size_t len,int behavior);asmlinkage long sys_process_madvise (int pidfd,const struct iovec __user *vec,size_t vlen,int behavior,unsigned int flags); +asmlinkage long sys_process_mrelease (int pidfd,unsigned int flags);asmlinkage long sys_remap_file_pages (unsigned long start,unsigned long size,unsigned long prot,unsigned long pgoff,unsigned long flags);
include/uapi/asm-generic/unistd.h +3 -1
@@ -877,9 +877,11 @@ __SYSCALL(__NR_landlock_restrict_self, sys_landlock_restrict_self)#define __NR_memfd_secret 447 __SYSCALL(__NR_memfd_secret, sys_memfd_secret)#endif +#define __NR_process_mrelease 448 +__SYSCALL(__NR_process_mrelease, sys_process_mrelease)#undef __NR_syscalls -#define __NR_syscalls 448 +#define __NR_syscalls 449/* * 32 bit systems traditionally used different
kernel/sys_ni.c +1 -0
@@ -289,6 +289,7 @@ COND_SYSCALL(munlockall);COND_SYSCALL (mincore);COND_SYSCALL (madvise);COND_SYSCALL (process_madvise); +COND_SYSCALL (process_mrelease);COND_SYSCALL (remap_file_pages);COND_SYSCALL (mbind);COND_SYSCALL_COMPAT (mbind);
Signature · 1 line
2.32.0.554.ge1b32706d8-goog
kernel test robot
build botRe: [PATCH v4 2/2] mm: wire up syscall process_mrelease
I love your patch! Perhaps something to improve:
Hi Suren,
I love your patch! Perhaps something to improve:
[auto build test WARNING on arm64/for-next/core]
[also build test WARNING on linus/master v5.14-rc4 next-20210802] [cannot apply to hnaz-linux-mm/master tip/x86/asm]
[If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Suren-Baghdasaryan/mm-introduce-process_mrelease-system-call/20210803-061733 base: https://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-next/core config: arm64-allyesconfig (attached as .config) compiler: aarch64-linux-gcc (GCC) 10.3.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/919f5f829cf9d1efd2df2935c7aae437a8c98a6d
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Suren-Baghdasaryan/mm-introduce-process_mrelease-system-call/20210803-061733 git checkout 919f5f829cf9d1efd2df2935c7aae437a8c98a6d
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-10.3.0 make.cross ARCH=arm64
If you fix the issue, kindly add following tag as appropriate
All warnings (new ones prefixed by >>):
| ^~~~~~~~~~~~
kernel/sys_ni.c:267:1: note: in expansion of macro 'COND_SYSCALL'
267 | COND_SYSCALL(keyctl);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_keyctl' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:268:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
268 | COND_SYSCALL_COMPAT(keyctl);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_landlock_create_ruleset' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:271:1: note: in expansion of macro 'COND_SYSCALL'
271 | COND_SYSCALL(landlock_create_ruleset);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_landlock_add_rule' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:272:1: note: in expansion of macro 'COND_SYSCALL'
272 | COND_SYSCALL(landlock_add_rule);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_landlock_restrict_self' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:273:1: note: in expansion of macro 'COND_SYSCALL'
273 | COND_SYSCALL(landlock_restrict_self);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_fadvise64_64' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:278:1: note: in expansion of macro 'COND_SYSCALL'
278 | COND_SYSCALL(fadvise64_64);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_swapon' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:281:1: note: in expansion of macro 'COND_SYSCALL'
281 | COND_SYSCALL(swapon);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_swapoff' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:282:1: note: in expansion of macro 'COND_SYSCALL'
282 | COND_SYSCALL(swapoff);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_mprotect' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:283:1: note: in expansion of macro 'COND_SYSCALL'
283 | COND_SYSCALL(mprotect);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_msync' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:284:1: note: in expansion of macro 'COND_SYSCALL'
284 | COND_SYSCALL(msync);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_mlock' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:285:1: note: in expansion of macro 'COND_SYSCALL'
285 | COND_SYSCALL(mlock);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_munlock' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:286:1: note: in expansion of macro 'COND_SYSCALL'
286 | COND_SYSCALL(munlock);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_mlockall' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:287:1: note: in expansion of macro 'COND_SYSCALL'
287 | COND_SYSCALL(mlockall);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_munlockall' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:288:1: note: in expansion of macro 'COND_SYSCALL'
288 | COND_SYSCALL(munlockall);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_mincore' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:289:1: note: in expansion of macro 'COND_SYSCALL'
289 | COND_SYSCALL(mincore);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_madvise' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:290:1: note: in expansion of macro 'COND_SYSCALL'
290 | COND_SYSCALL(madvise);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_process_madvise' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:291:1: note: in expansion of macro 'COND_SYSCALL'
291 | COND_SYSCALL(process_madvise);
| ^~~~~~~~~~~~
>> arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_process_mrelease' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:292:1: note: in expansion of macro 'COND_SYSCALL'
292 | COND_SYSCALL(process_mrelease);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_remap_file_pages' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:293:1: note: in expansion of macro 'COND_SYSCALL'
293 | COND_SYSCALL(remap_file_pages);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_mbind' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:294:1: note: in expansion of macro 'COND_SYSCALL'
294 | COND_SYSCALL(mbind);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_mbind' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:295:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
295 | COND_SYSCALL_COMPAT(mbind);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_get_mempolicy' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:296:1: note: in expansion of macro 'COND_SYSCALL'
296 | COND_SYSCALL(get_mempolicy);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_get_mempolicy' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:297:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
297 | COND_SYSCALL_COMPAT(get_mempolicy);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_set_mempolicy' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:298:1: note: in expansion of macro 'COND_SYSCALL'
298 | COND_SYSCALL(set_mempolicy);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_set_mempolicy' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:299:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
299 | COND_SYSCALL_COMPAT(set_mempolicy);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_migrate_pages' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:300:1: note: in expansion of macro 'COND_SYSCALL'
300 | COND_SYSCALL(migrate_pages);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_migrate_pages' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:301:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
301 | COND_SYSCALL_COMPAT(migrate_pages);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_move_pages' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:302:1: note: in expansion of macro 'COND_SYSCALL'
302 | COND_SYSCALL(move_pages);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_move_pages' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:303:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
303 | COND_SYSCALL_COMPAT(move_pages);
| ^~~~~~~~~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_perf_event_open' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:305:1: note: in expansion of macro 'COND_SYSCALL'
305 | COND_SYSCALL(perf_event_open);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_accept4' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:306:1: note: in expansion of macro 'COND_SYSCALL'
306 | COND_SYSCALL(accept4);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_recvmmsg' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:307:1: note: in expansion of macro 'COND_SYSCALL'
307 | COND_SYSCALL(recvmmsg);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:76:25: warning: no previous prototype for '__arm64_sys_recvmmsg_time32' [-Wmissing-prototypes]
76 | asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~
kernel/sys_ni.c:308:1: note: in expansion of macro 'COND_SYSCALL'
308 | COND_SYSCALL(recvmmsg_time32);
| ^~~~~~~~~~~~
arch/arm64/include/asm/syscall_wrapper.h:41:25: warning: no previous prototype for '__arm64_compat_sys_recvmmsg_time32' [-Wmissing-prototypes]
41 | asmlinkage long __weak __arm64_compat_sys_##name(const struct pt_regs *regs) \
| ^~~~~~~~~~~~~~~~~~~
kernel/sys_ni.c:309:1: note: in expansion of macro 'COND_SYSCALL_COMPAT'
309 | COND_SYSCALL_COMPAT(recvmmsg_time32);vim +/__arm64_sys_process_mrelease +76 arch/arm64/include/asm/syscall_wrapper.h
4378a7d4be30ec Mark Rutland 2018-07-11 50
4378a7d4be30ec Mark Rutland 2018-07-11 51 #define __SYSCALL_DEFINEx(x, name, ...) \
4378a7d4be30ec Mark Rutland 2018-07-11 52 asmlinkage long __arm64_sys##name(const struct pt_regs *regs); \
4378a7d4be30ec Mark Rutland 2018-07-11 53 ALLOW_ERROR_INJECTION(__arm64_sys##name, ERRNO); \
4378a7d4be30ec Mark Rutland 2018-07-11 54 static long __se_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__)); \
4378a7d4be30ec Mark Rutland 2018-07-11 55 static inline long __do_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__)); \
4378a7d4be30ec Mark Rutland 2018-07-11 56 asmlinkage long __arm64_sys##name(const struct pt_regs *regs) \
4378a7d4be30ec Mark Rutland 2018-07-11 57 { \
4378a7d4be30ec Mark Rutland 2018-07-11 58 return __se_sys##name(SC_ARM64_REGS_TO_ARGS(x,__VA_ARGS__)); \
4378a7d4be30ec Mark Rutland 2018-07-11 59 } \
4378a7d4be30ec Mark Rutland 2018-07-11 60 static long __se_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__)) \
4378a7d4be30ec Mark Rutland 2018-07-11 61 { \
4378a7d4be30ec Mark Rutland 2018-07-11 62 long ret = __do_sys##name(__MAP(x,__SC_CAST,__VA_ARGS__)); \
4378a7d4be30ec Mark Rutland 2018-07-11 63 __MAP(x,__SC_TEST,__VA_ARGS__); \
4378a7d4be30ec Mark Rutland 2018-07-11 64 __PROTECT(x, ret,__MAP(x,__SC_ARGS,__VA_ARGS__)); \
4378a7d4be30ec Mark Rutland 2018-07-11 65 return ret; \
4378a7d4be30ec Mark Rutland 2018-07-11 66 } \
4378a7d4be30ec Mark Rutland 2018-07-11 67 static inline long __do_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))
4378a7d4be30ec Mark Rutland 2018-07-11 68
4378a7d4be30ec Mark Rutland 2018-07-11 69 #define SYSCALL_DEFINE0(sname) \
4378a7d4be30ec Mark Rutland 2018-07-11 70 SYSCALL_METADATA(_##sname, 0); \
0e358bd7b7ebd2 Sami Tolvanen 2019-05-24 71 asmlinkage long __arm64_sys_##sname(const struct pt_regs *__unused); \
4378a7d4be30ec Mark Rutland 2018-07-11 72 ALLOW_ERROR_INJECTION(__arm64_sys_##sname, ERRNO); \
0e358bd7b7ebd2 Sami Tolvanen 2019-05-24 73 asmlinkage long __arm64_sys_##sname(const struct pt_regs *__unused)
4378a7d4be30ec Mark Rutland 2018-07-11 74
c27eccfe4d6c74 Sami Tolvanen 2019-09-10 75 #define COND_SYSCALL(name) \
c27eccfe4d6c74 Sami Tolvanen 2019-09-10 @76 asmlinkage long __weak __arm64_sys_##name(const struct pt_regs *regs) \
c27eccfe4d6c74 Sami Tolvanen 2019-09-10 77 { \
c27eccfe4d6c74 Sami Tolvanen 2019-09-10 78 return sys_ni_syscall(); \
c27eccfe4d6c74 Sami Tolvanen 2019-09-10 79 }4378a7d4be30ec Mark Rutland 2018-07-11 80
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
[...]
Quoted 2 lines
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extending
I have no idea how stable these links are. Referencing via message id is the common practice. For this link, we'd use
https://lkml.kernel.org/r/[email protected]/
instead.
Quoted 7 lines
pidfd_send_signalto allow memory reaping usingoom_reaperthread;
- https://lore.kernel.org/patchwork/patch/1338196 extendingpidfd_send_signalto reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to addMADV_DONTNEEDsupport forprocess_madviseimplementing synchronous memory reaping.
Quoted 10 lines
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
The reasoning was that the new variant ofprocess_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest ofprocess_madvise
From the userspace point of view it was awkward and inconvenient to provide
memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.
I'd condense this description and only reference previous discussions to put a main focus on what this patch actually does. Like
"
After previous discussions [1, 2, 3] the decision was made to introduce a dedicated system call to cover this use case.
...
[1] https://lkml.kernel.org/r/[email protected]/
"
Quoted 23 lines
The API is as follows,
int
process_mrelease(int pidfd, unsigned int flags);DESCRIPTION
Theprocess_mrelease()system call is used to free the memory of a process which was sent a SIGKILL signal.The pidfd selects the process referred to by the PID file descriptor.
(Seepidofd_open(2)for further information)The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success,process_mrelease()returns 0. On error, -1 is returned and errno is set to indicate the error.ERRORS
EBADF pidfd is not a valid PID file descriptor.
Quoted 8 lines
EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
EINVAL flags is not 0.
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.
Hm, I do wonder if it would make sense to have a mode (e.g., via a flag) to reap all but shared memory from a dying process. Future work.
Quoted 3 lines
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).
Maybe "This system call is not supported, for example, without MMU support built into Linux."
Quoted 30 lines
ESRCH The target process does not exist (i.e., it has terminated
and been waited on).Signed-off-by Suren Baghdasaryan <[email protected]>---
changes in v4:
- Replacedmmap_read_lock()withmmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentationmm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)mm/oom_kill.c +2
-0@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h> @@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } +
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1155
Quoted 10 lines
mm/oom_kill.c +10
-0+SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ; +struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 )
if (flags)
Quoted 23 lines
mm/oom_kill.c +23
-0+return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID ); +if (!task) { + ret = -ESRCH ; + goto put_pid; + } + +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task); +if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1188
Quoted 10 lines
mm/oom_kill.c +10
-0+ ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ;
I'm not an expert on __oom_reap_task_mm(), but the whole approach makes sense to. So feel free to add my
Signature · 3 lines
Thanks, David / dhildenb
Quoted 6 lines · David Hildenbrand
[...]
Quoted 2 lines
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extending
Quoted 6 lines
I have no idea how stable these links are. Referencing via message id is the common practice. For this link, we'd use
https://lkml.kernel.org/r/[email protected]/
instead.
Ack.
Quoted 22 lines
Quoted 8 lines
pidfd_send_signalto allow memory reaping usingoom_reaperthread;
- https://lore.kernel.org/patchwork/patch/1338196 extendingpidfd_send_signalto reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to addMADV_DONTNEEDsupport forprocess_madviseimplementing synchronous memory reaping.Quoted 10 lines
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
The reasoning was that the new variant ofprocess_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest ofprocess_madvise
From the userspace point of view it was awkward and inconvenient to provide
memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.I'd condense this description and only reference previous discussions to put a main focus on what this patch actually does. Like
Quoted 9 lines
"
After previous discussions [1, 2, 3] the decision was made to introduce a dedicated system call to cover this use case....
Ack.
Quoted 27 lines
Quoted 27 lines
The API is as follows,
int
process_mrelease(int pidfd, unsigned int flags);DESCRIPTION
Theprocess_mrelease()system call is used to free the memory of a process which was sent a SIGKILL signal.The pidfd selects the process referred to by the PID file descriptor.
(Seepidofd_open(2)for further information)The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success,process_mrelease()returns 0. On error, -1 is returned and errno is set to indicate the error.ERRORS
EBADF pidfd is not a valid PID file descriptor.EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
Quoted 7 lines
Quoted 4 lines
EINVAL flags is not 0.
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.Hm, I do wonder if it would make sense to have a mode (e.g., via a flag) to reap all but shared memory from a dying process. Future work.
Agree. Let's keep it simple for now and will expand when the need arises.
Quoted 7 lines
Quoted 4 lines
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).Maybe "This system call is not supported, for example, without MMU support built into Linux."
Ack.
Quoted 35 lines
Quoted 35 lines
ESRCH The target process does not exist (i.e., it has terminated
and been waited on).Signed-off-by Suren Baghdasaryan <[email protected]>---
changes in v4:
- Replacedmmap_read_lock()withmmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentationmm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)mm/oom_kill.c +6
-0@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h> @@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } + +SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ;
Quoted 8 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1155
Quoted 6 lines
mm/oom_kill.c +6
-0+struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 )if (flags)
Ack.
Quoted 38 lines
Quoted 24 lines
mm/oom_kill.c +23
-0+return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID ); +if (!task) { + ret = -ESRCH ; + goto put_pid; + } + +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task); +if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {mm/oom_kill.c void pagefault_out_of_memory(void) · line 1188
Quoted 10 lines
mm/oom_kill.c +10
-0+ ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ;I'm not an expert on
__oom_reap_task_mm(), but the whole approach makes sense to. So feel free to add my
Quoted 1 line
Thanks! I see Michal also asked for some documentation changes and a simple code change, so I won't roll your Acked-by automatically into the next version but would appreciate it on the final version :) Will post the next rev later today or tomorrow morning. Thanks for the review!
Quoted 6 lines
--
Thanks,David / dhildenb
Michal Hocko
maintainerQuoted 19 lines · Suren Baghdasaryan
In modern systems it's not unusual to have a system component monitoring memory conditions of the system and tasked with keeping system memory pressure under control. One way to accomplish that is to kill non-essential processes to free up memory for more important ones. Examples of this are Facebook's OOM killer daemon called oomd and Android's low memory killer daemon called lmkd.
For such system component it's important to be able to free memory quickly and efficiently. Unfortunately the time process takes to free up its memory after receiving a SIGKILL might vary based on the state of the process (uninterruptible sleep), size and OPP level of the core the process is running. A mechanism to free resources of the target process in a more predictable way would improve system's ability to control its memory pressure.
Introduceprocess_mreleasesystem call that releases memory of a dying process from the context of the caller. This way the memory is freed in a more controllable way with CPU affinity and priority of the caller. The workload of freeing the memory will also be charged to the caller. The operation is allowed only on a dying process.
Quoted 2 lines
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extending
Please use the msg-id based urls https://lore.kernel.org/lkml/[email protected]/
Quoted 2 lines
pidfd_send_signalto allow memory reaping usingoom_reaperthread;
- https://lore.kernel.org/patchwork/patch/1338196 extending
https://lore.kernel.org/linux-api/[email protected]/
Quoted 4 lines
pidfd_send_signalto reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to addMADV_DONTNEEDsupport forprocess_madviseimplementing synchronous memory reaping.
https://lore.kernel.org/linux-api/[email protected]/
Quoted 2 lines
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
https://lore.kernel.org/linux-api/[email protected]/
Quoted 9 lines
The reasoning was that the new variant of
process_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest ofprocess_madviseQuoted 1 line
From the userspace point of view it was awkward and inconvenient to provide
memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.
Quoted 7 lines
The API is as follows,
int
process_mrelease(int pidfd, unsigned int flags);DESCRIPTION
Theprocess_mrelease()system call is used to free the memory of a process which was sent a SIGKILL signal.
This is not really precise. The implementation will allow to use the syscall on any exiting or fatal signal received process. Not just those that have been SIGKILLed, right? For the purpose of the man page I would go with exiting process for the wording.
Quoted 14 lines
The pidfd selects the process referred to by the PID file descriptor.
(Seepidofd_open(2)for further information)The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success,process_mrelease()returns 0. On error, -1 is returned and errno is set to indicate the error.ERRORS
EBADF pidfd is not a valid PID file descriptor.
Quoted 8 lines
EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
EINVAL flags is not 0.
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.
again, wording here. I would go with
EINVAL The memory of the task cannot be released because the
process is not exiting, the address space is shared with an alive process or there is a core dump is in progress..
Quoted 50 lines
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).ESRCH The target process does not exist (i.e., it has terminated
and been waited on).Signed-off-by Suren Baghdasaryan <[email protected]>---
changes in v4:
- Replacedmmap_read_lock()withmmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentationmm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)mm/oom_kill.c +19
-0@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h> @@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } + +SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ; +struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 ) +return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID );
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1172
Quoted 10 lines
mm/oom_kill.c +10
-0+if (!task) { + ret = -ESRCH ; + goto put_pid; + } + +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task);
Don't we need find_lock_task_mm here?
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1178
Quoted 6 lines
mm/oom_kill.c +6
-0+if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {
Do we want to treat MMF_OOM_SKIP as a failure?
Quoted 25 lines
mm/oom_kill.c +23
-0+ ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ; +mmap_read_unlock (mm); + +put_mm: +mmput (mm); +put_task: +put_task_struct (task); +put_pid: +put_pid (pid); +return ret; +#else +return -ENOSYS ; +#endif /* CONFIG_MMU */ +}Signature · 1 line
2.32.0.554.ge1b32706d8-goog
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 21 lines · Michal Hocko
Quoted 19 lines · Suren Baghdasaryan
In modern systems it's not unusual to have a system component monitoring memory conditions of the system and tasked with keeping system memory pressure under control. One way to accomplish that is to kill non-essential processes to free up memory for more important ones. Examples of this are Facebook's OOM killer daemon called oomd and Android's low memory killer daemon called lmkd.
For such system component it's important to be able to free memory quickly and efficiently. Unfortunately the time process takes to free up its memory after receiving a SIGKILL might vary based on the state of the process (uninterruptible sleep), size and OPP level of the core the process is running. A mechanism to free resources of the target process in a more predictable way would improve system's ability to control its memory pressure.
Introduceprocess_mreleasesystem call that releases memory of a dying process from the context of the caller. This way the memory is freed in a more controllable way with CPU affinity and priority of the caller. The workload of freeing the memory will also be charged to the caller. The operation is allowed only on a dying process.
Quoted 4 lines
Quoted 2 lines
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extendingPlease use the msg-id based urls https://lore.kernel.org/lkml/[email protected]/
Will do. Thanks!
Quoted 31 lines
Quoted 3 lines
pidfd_send_signalto allow memory reaping usingoom_reaperthread;
- https://lore.kernel.org/patchwork/patch/1338196 extendinghttps://lore.kernel.org/linux-api/[email protected]/
Quoted 4 lines
pidfd_send_signalto reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to addMADV_DONTNEEDsupport forprocess_madviseimplementing synchronous memory reaping.https://lore.kernel.org/linux-api/[email protected]/
Quoted 2 lines
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
https://lore.kernel.org/linux-api/[email protected]/
Quoted 13 lines
The reasoning was that the new variant of
process_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest ofprocess_madviseQuoted 1 line
From the userspace point of view it was awkward and inconvenient to provide
memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.
The API is as follows,
int process_mrelease(int pidfd, unsigned int flags);
Quoted 8 lines
Quoted 3 lines
DESCRIPTION
Theprocess_mrelease()system call is used to free the memory of a process which was sent a SIGKILL signal.This is not really precise. The implementation will allow to use the syscall on any exiting or fatal signal received process. Not just those that have been SIGKILLed, right? For the purpose of the man page I would go with exiting process for the wording.
Ack.
Quoted 21 lines
Quoted 21 lines
The pidfd selects the process referred to by the PID file descriptor.
(Seepidofd_open(2)for further information)The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success,process_mrelease()returns 0. On error, -1 is returned and errno is set to indicate the error.ERRORS
EBADF pidfd is not a valid PID file descriptor.EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
EINVAL flags is not 0.
Quoted 8 lines
Quoted 2 lines
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.again, wording here. I would go with
EINVAL The memory of the task cannot be released because the
process is not exiting, the address space is shared with an alive process or there is a core dump is in progress..
Ack.
Quoted 54 lines
Quoted 54 lines
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).ESRCH The target process does not exist (i.e., it has terminated
and been waited on).Signed-off-by Suren Baghdasaryan <[email protected]>---
changes in v4:
- Replacedmmap_read_lock()withmmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentationmm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)mm/oom_kill.c +23
-0@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h> @@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } + +SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ; +struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 ) +return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID ); +if (!task) { + ret = -ESRCH ; + goto put_pid; + }
Quoted 8 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1172
Quoted 6 lines
mm/oom_kill.c +6
-0+ +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task);Don't we need
find_lock_task_mmhere?
Yes, we do. Will fix in the next rev.
Quoted 9 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1178
Quoted 7 lines
mm/oom_kill.c +6
-0+if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {Do we want to treat
MMF_OOM_SKIPas a failure?
Yeah, I don't think we want to create additional contention if oom-killer is already working on this mm. Should we return EBUSY in this case? Other possible options is ESRCH, indicating that this process is a goner, so don't bother. WDYT?
Quoted 18 lines
Quoted 18 lines
mm/oom_kill.c +17
-0+ ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ; +mmap_read_unlock (mm); + +put_mm: +mmput (mm); +put_task: +put_task_struct (task); +put_pid:
Quoted 9 lines
Quoted 8 lines
mm/oom_kill.c +6
-1+put_pid (pid); +return ret; +#else +return -ENOSYS ; +#endif /* CONFIG_MMU */ +} --2.32.0.554.ge1b32706d8-goog
Thanks for the review!
Quoted 3 lines
--
Michal Hocko
SUSE Labs
Quoted 177 lines · Suren Baghdasaryan
Quoted 21 lines · Michal Hocko
Quoted 19 lines · Suren Baghdasaryan
In modern systems it's not unusual to have a system component monitoring memory conditions of the system and tasked with keeping system memory pressure under control. One way to accomplish that is to kill non-essential processes to free up memory for more important ones. Examples of this are Facebook's OOM killer daemon called oomd and Android's low memory killer daemon called lmkd.
For such system component it's important to be able to free memory quickly and efficiently. Unfortunately the time process takes to free up its memory after receiving a SIGKILL might vary based on the state of the process (uninterruptible sleep), size and OPP level of the core the process is running. A mechanism to free resources of the target process in a more predictable way would improve system's ability to control its memory pressure.
Introduceprocess_mreleasesystem call that releases memory of a dying process from the context of the caller. This way the memory is freed in a more controllable way with CPU affinity and priority of the caller. The workload of freeing the memory will also be charged to the caller. The operation is allowed only on a dying process.Quoted 4 lines
Quoted 2 lines
Previously I proposed a number of alternatives to accomplish this:
- https://lore.kernel.org/patchwork/patch/1060407 extendingPlease use the msg-id based urls https://lore.kernel.org/lkml/[email protected]/
Will do. Thanks!
Quoted 31 lines
Quoted 3 lines
pidfd_send_signalto allow memory reaping usingoom_reaperthread;
- https://lore.kernel.org/patchwork/patch/1338196 extendinghttps://lore.kernel.org/linux-api/[email protected]/
Quoted 4 lines
pidfd_send_signalto reap memory of the target process synchronously from the context of the caller;
- https://lore.kernel.org/patchwork/patch/1344419/ to addMADV_DONTNEEDsupport forprocess_madviseimplementing synchronous memory reaping.https://lore.kernel.org/linux-api/[email protected]/
Quoted 2 lines
The end of the last discussion culminated with suggestion to introduce a dedicated system call (https://lore.kernel.org/patchwork/patch/1344418/#1553875)
https://lore.kernel.org/linux-api/[email protected]/
Quoted 13 lines
The reasoning was that the new variant of
process_madvise
a) does not work on an address range
b) is destructive
c) doesn't share much code at all with the rest ofprocess_madviseQuoted 1 line
From the userspace point of view it was awkward and inconvenient to provide
memory range for this operation that operates on the entire address space. Using special flags or address values to specify the entire address space was too hacky.
The API is as follows,
int process_mrelease(int pidfd, unsigned int flags);Quoted 8 lines
Quoted 3 lines
DESCRIPTION
Theprocess_mrelease()system call is used to free the memory of a process which was sent a SIGKILL signal.This is not really precise. The implementation will allow to use the syscall on any exiting or fatal signal received process. Not just those that have been SIGKILLed, right? For the purpose of the man page I would go with exiting process for the wording.
Ack.
Quoted 21 lines
Quoted 21 lines
The pidfd selects the process referred to by the PID file descriptor.
(Seepidofd_open(2)for further information)The flags argument is reserved for future use; currently, this argument must be specified as 0.
RETURN VALUE
On success,process_mrelease()returns 0. On error, -1 is returned and errno is set to indicate the error.ERRORS
EBADF pidfd is not a valid PID file descriptor.EAGAIN Failed to release part of the address space.
EINTR The call was interrupted by a signal; see signal(7).
EINVAL flags is not 0.
Quoted 8 lines
Quoted 2 lines
EINVAL The task does not have a pending SIGKILL or its memory is
shared with another process with no pending SIGKILL.again, wording here. I would go with
EINVAL The memory of the task cannot be released because the
process is not exiting, the address space is shared with an alive process or there is a core dump is in progress..Ack.
Quoted 54 lines
Quoted 54 lines
ENOSYS This system call is not supported by kernels built with no
MMU support (CONFIG_MMU=n).ESRCH The target process does not exist (i.e., it has terminated
and been waited on).Signed-off-by Suren Baghdasaryan <[email protected]>---
changes in v4:
- Replacedmmap_read_lock()withmmap_read_lock_killable(), per Michal Hocko
- Added EINTR error in the manual pages documentationmm/oom_kill.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+)mm/oom_kill.c +23
-0@@ -28,6 +28,7 @@#include <linux/sched/task.h> #include <linux/sched/debug.h> #include <linux/swap.h> +#include <linux/syscalls.h> #include <linux/timex.h> #include <linux/jiffies.h> #include <linux/cpuset.h> @@ -1141,3 +1142,60 @@ void pagefault_out_of_memory(void)out_of_memory (&oc);mutex_unlock (&oom_lock); } + +SYSCALL_DEFINE2(process_mrelease, int, pidfd, unsigned int, flags) +{ +#ifdef CONFIG_MMU +struct mm_struct *mm =NULL ; +struct task_struct *task; +unsigned int f_flags; +struct pid *pid; +long ret =0 ; + +if (flags !=0 ) +return -EINVAL ; + + pid =pidfd_get_pid (pidfd, &f_flags); +if (IS_ERR (pid)) +return PTR_ERR (pid); + + task =get_pid_task (pid,PIDTYPE_PID ); +if (!task) { + ret = -ESRCH ; + goto put_pid; + }Quoted 8 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1172
Quoted 6 lines
mm/oom_kill.c +6
-0+ +/* +* If the task is dying and in the process of releasing its memory +* then get its mm. +*/ +task_lock (task);Don't we need
find_lock_task_mmhere?Yes, we do. Will fix in the next rev.
Quoted 8 lines
Quoted 7 lines
mm/oom_kill.c +6
-0+if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {
Quoted 6 lines
Quoted 1 line
Do we want to treat
MMF_OOM_SKIPas a failure?Yeah, I don't think we want to create additional contention if oom-killer is already working on this mm. Should we return EBUSY in this case? Other possible options is ESRCH, indicating that this process is a goner, so don't bother. WDYT?
After considering this some more I think ESRCH would be more appropriate. EBUSY might be understood as "I need to retry at a better time", which is not what we want here.
I posted v5 at https://lore.kernel.org/patchwork/patch/1471926 with suggested changes.
Thanks,
Suren.
Quoted 34 lines
Quoted 19 lines
Quoted 19 lines
mm/oom_kill.c +17
-0+ ret = -EINVAL ; + goto put_task; + } + +if (mmap_read_lock_killable (mm)) { + ret = -EINTR ; + goto put_mm; + } +if (!__oom_reap_task_mm (mm)) + ret = -EAGAIN ; +mmap_read_unlock (mm); + +put_mm: +mmput (mm); +put_task: +put_task_struct (task); +put_pid:Quoted 9 lines
Quoted 8 lines
mm/oom_kill.c +6
-1+put_pid (pid); +return ret; +#else +return -ENOSYS ; +#endif /* CONFIG_MMU */ +} --2.32.0.554.ge1b32706d8-goog
Thanks for the review!
Quoted 3 lines
--
Michal Hocko
SUSE Labs
Michal Hocko
maintainerQuoted 1 line · Suren Baghdasaryan
On Tue, Aug 3, 2021 at 10:27 AM Suren Baghdasaryan <[email protected]> wrote:
[...]
Quoted 9 lines
Quoted 9 lines
Quoted 8 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1178
Quoted 6 lines
mm/oom_kill.c +6
-0+if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {Do we want to treat
MMF_OOM_SKIPas a failure?
Quoted 8 lines
Quoted 4 lines
Yeah, I don't think we want to create additional contention if oom-killer is already working on this mm. Should we return EBUSY in this case? Other possible options is ESRCH, indicating that this process is a goner, so don't bother. WDYT?
After considering this some more I think ESRCH would be more appropriate. EBUSY might be understood as "I need to retry at a better time", which is not what we want here.
Why cannot we simply return 0 in that case. The work has been done already by the kernel so why should we tell the caller that there was something wrong?
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 18 lines · Michal Hocko
Quoted 1 line · Suren Baghdasaryan
On Tue, Aug 3, 2021 at 10:27 AM Suren Baghdasaryan <[email protected]> wrote:
[...]
Quoted 14 lines
Quoted 13 lines
Quoted 8 lines
mm/oom_kill.c void pagefault_out_of_memory(void) · line 1178
Quoted 6 lines
mm/oom_kill.c +6
-0+if (task_will_free_mem (task) && (task->flags &PF_KTHREAD ) ==0 ) { + mm = task->mm; +mmget (mm); + } +task_unlock (task); +if (!mm) {Do we want to treat
MMF_OOM_SKIPas a failure?Yeah, I don't think we want to create additional contention if oom-killer is already working on this mm. Should we return EBUSY in this case? Other possible options is ESRCH, indicating that this process is a goner, so don't bother. WDYT?
Quoted 7 lines
Quoted 3 lines
After considering this some more I think ESRCH would be more appropriate. EBUSY might be understood as "I need to retry at a better time", which is not what we want here.
Why cannot we simply return 0 in that case. The work has been done already by the kernel so why should we tell the caller that there was something wrong?
Ah, you are right. I was under the impression that MMF_OOM_SKIP means oom-killer is reaping the mm, but looks like it means that mm was already reaped. If that's true then I agree, returning 0 is the right move here. Will fix.
Quoted 4 lines
--
Michal Hocko
SUSE Labs