diff mbox series

[2/2] mm: support vector address ranges for process_madvise

Message ID 20200423145215.72666-2-minchan@kernel.org (mailing list archive)
State New, archived
Headers show
Series [1/2] fs: use current->mm for io_uring | expand

Commit Message

Minchan Kim April 23, 2020, 2:52 p.m. UTC
This patch extends a) process_madvise(2) support vector address ranges
in a system call and then b) support the vector address ranges to
local process as well as external process.

Android app has thousands of vmas due to zygote so it's totally waste
of CPU and power if we should call the syscall one by one for each vma.
(With testing 2000-vma syscall vs 1-vector syscall, it showed 15%
performance improvement. I think it would be bigger in real practice
because the testing ran very cache friendly environment).

Another potential use case for the vector range is to amortize the
cost of TLB shootdowns for multiple ranges when using MADV_DONTNEED;
this could benefit users like TCP receive zerocopy and malloc
implementations. In future, we could find more usecases for other
advises so let's make it happens as API since we introduce a new syscall
at this moment. With that, existing madvise(2) user could replace it with
process_madvise(2) with their own pid if they want to have batch address
ranges support feature.

So finally, the API is as follows,

  ssize_t process_madvise(idtype_t idtype, id_t id,
		const struct iovec *iovec, unsigned long vlen,
                int advice, unsigned long flags);

DESCRIPTION
  The process_madvise() system call is used to give advice or directions
  to the kernel about the address ranges from external process as well as
  local process. It provides the advice to address ranges of process
  described by iovec and vlen. The goal of such advice is to improve system
  or application performance.

  The idtype and id arguments select the target process to be advised as
  follows:

    idtype == P_PID
      select the process whose process ID matches id

    idtype == P_PIDFD
      select the process referred to by the PID file descriptor
      specified in id. (See pidofd_open(2) for further information)

  The pointer iovec points to an array of iovec structures, defined in
  <sys/uio.h> as:

    struct iovec {
    	void *iov_base;		/* starting address */
	size_t iov_len;		/* number of bytes to be advised */
    };

  The iovec describes address ranges beginning at address(iov_base)
  and with size length of bytes(iov_len).

  The vlen represents the number of elements in iovec.

  The advice is indicated in the advice argument, which is one of the
  following at this moment if the target process specified by idtype and
  id is external.

    MADV_COLD
    MADV_PAGEOUT
    MADV_MERGEABLE
    MADV_UNMERGEABLE

  Permission to provide a hint to external process is governed by a
  ptrace access mode PTRACE_MODE_ATTACH_FSCREDS check; see ptrace(2).

  The process_madvise supports every advice madvise(2) has if target
  process is in same thread group with calling process so user could
  use process_madvise(2) to extend existing madvise(2) to support
  vector address ranges.

RETURN VALUE
  On success, process_madvise() returns the number of bytes advised.
  This return value may be less than the total number of requested
  bytes, if an error occurred. The caller should check return value
  to determine whether a partial advice occurred.

Cc: David Rientjes <rientjes@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Tim Murray <timmurray@google.com>
Cc: Daniel Colascione <dancol@google.com>
Cc: Sonny Rao <sonnyrao@google.com>
Cc: Brian Geffon <bgeffon@google.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: John Dias <joaodias@google.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: SeongJae Park <sj38.park@gmail.com>
Cc: Oleksandr Natalenko <oleksandr@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Sandeep Patil <sspatil@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: <linux-man@vger.kernel.org>
Signed-off-by: Minchan Kim <minchan@kernel.org>
---
 mm/madvise.c | 47 ++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 40 insertions(+), 7 deletions(-)

Comments

Matthew Wilcox (Oracle) April 23, 2020, 3:14 p.m. UTC | #1
On Thu, Apr 23, 2020 at 07:52:15AM -0700, Minchan Kim wrote:
> +SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
> +		const struct iovec __user *, vec, unsigned long, vlen,
> +		int, behavior, unsigned long, flags)
> +{

Don't we now need a compat version of this that calls compat_import_iovec()?
Minchan Kim April 23, 2020, 7:58 p.m. UTC | #2
On Thu, Apr 23, 2020 at 08:14:10AM -0700, Matthew Wilcox wrote:
> On Thu, Apr 23, 2020 at 07:52:15AM -0700, Minchan Kim wrote:
> > +SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
> > +		const struct iovec __user *, vec, unsigned long, vlen,
> > +		int, behavior, unsigned long, flags)
> > +{
> 
> Don't we now need a compat version of this that calls compat_import_iovec()?

Yub, Thanks.

From 72fd6dbbab734803110817ef180ac2c6e3c2ca2a Mon Sep 17 00:00:00 2001
From: Minchan Kim <minchan@kernel.org>
Date: Thu, 23 Apr 2020 12:53:53 -0700
Subject: [PATCH] mm: support compat_sys_proess_madvise

This patch supports compat syscall for process_madvise

Signed-off-by: Minchan Kim <minchan@kernel.org>
---
 arch/arm64/include/asm/unistd32.h         |  2 +-
 arch/mips/kernel/syscalls/syscall_n32.tbl |  2 +-
 arch/parisc/kernel/syscalls/syscall.tbl   |  2 +-
 arch/powerpc/kernel/syscalls/syscall.tbl  |  2 +-
 arch/s390/kernel/syscalls/syscall.tbl     |  2 +-
 arch/sparc/kernel/syscalls/syscall.tbl    |  2 +-
 arch/x86/entry/syscalls/syscall_32.tbl    |  2 +-
 arch/x86/entry/syscalls/syscall_64.tbl    |  3 +-
 include/linux/compat.h                    |  3 ++
 include/uapi/asm-generic/unistd.h         |  3 +-
 kernel/sys_ni.c                           |  1 +
 mm/madvise.c                              | 60 +++++++++++++++++------
 12 files changed, 60 insertions(+), 24 deletions(-)

diff --git a/arch/arm64/include/asm/unistd32.h b/arch/arm64/include/asm/unistd32.h
index a1eec8d879d4..4958633ea7c2 100644
--- a/arch/arm64/include/asm/unistd32.h
+++ b/arch/arm64/include/asm/unistd32.h
@@ -884,7 +884,7 @@ __SYSCALL(__NR_openat2, sys_openat2)
 #define __NR_pidfd_getfd 438
 __SYSCALL(__NR_pidfd_getfd, sys_pidfd_getfd)
 #define __NR_process_madvise 439
-__SYSCALL(__NR_process_madvise, sys_process_madvise)
+__SYSCALL(__NR_process_madvise, compat_sys_process_madvise)
 
 /*
  * Please add new compat syscalls above this comment and update
diff --git a/arch/mips/kernel/syscalls/syscall_n32.tbl b/arch/mips/kernel/syscalls/syscall_n32.tbl
index 8ec8c558aa9c..1a20c9fcdcdb 100644
--- a/arch/mips/kernel/syscalls/syscall_n32.tbl
+++ b/arch/mips/kernel/syscalls/syscall_n32.tbl
@@ -376,4 +376,4 @@
 435	n32	clone3				__sys_clone3
 437	n32	openat2				sys_openat2
 438	n32	pidfd_getfd			sys_pidfd_getfd
-439	n32	process_madvise			sys_process_madvise
+439	n32	process_madvise			compat_sys_process_madvise
diff --git a/arch/parisc/kernel/syscalls/syscall.tbl b/arch/parisc/kernel/syscalls/syscall.tbl
index 09c3b5dc6855..3cbf5f5edab5 100644
--- a/arch/parisc/kernel/syscalls/syscall.tbl
+++ b/arch/parisc/kernel/syscalls/syscall.tbl
@@ -435,4 +435,4 @@
 435	common	clone3				sys_clone3_wrapper
 437	common	openat2				sys_openat2
 438	common	pidfd_getfd			sys_pidfd_getfd
-439	common	process_madvise			sys_process_madvise
+439	common	process_madvise			sys_process_madvise		compat_sys_process_madvise
diff --git a/arch/powerpc/kernel/syscalls/syscall.tbl b/arch/powerpc/kernel/syscalls/syscall.tbl
index 8a0c7239a6f3..03b8bc7707db 100644
--- a/arch/powerpc/kernel/syscalls/syscall.tbl
+++ b/arch/powerpc/kernel/syscalls/syscall.tbl
@@ -527,4 +527,4 @@
 435	spu	clone3				sys_ni_syscall
 437	common	openat2				sys_openat2
 438	common	pidfd_getfd			sys_pidfd_getfd
-439	common	process_madvise			sys_process_madvise
+439	common	process_madvise			sys_process_madvise		compat_sys_process_madvise
diff --git a/arch/s390/kernel/syscalls/syscall.tbl b/arch/s390/kernel/syscalls/syscall.tbl
index 8dc8bfd958ea..69f4a5459f0e 100644
--- a/arch/s390/kernel/syscalls/syscall.tbl
+++ b/arch/s390/kernel/syscalls/syscall.tbl
@@ -440,4 +440,4 @@
 435  common	clone3			sys_clone3			sys_clone3
 437  common	openat2			sys_openat2			sys_openat2
 438  common	pidfd_getfd		sys_pidfd_getfd			sys_pidfd_getfd
-439  common	process_madvise		sys_process_madvise		sys_process_madvise
+439  common	process_madvise		sys_process_madvise		compat_sys_process_madvise
diff --git a/arch/sparc/kernel/syscalls/syscall.tbl b/arch/sparc/kernel/syscalls/syscall.tbl
index 6f6e66dd51f9..bff349fcba0d 100644
--- a/arch/sparc/kernel/syscalls/syscall.tbl
+++ b/arch/sparc/kernel/syscalls/syscall.tbl
@@ -483,4 +483,4 @@
 # 435 reserved for clone3
 437	common	openat2			sys_openat2
 438	common	pidfd_getfd			sys_pidfd_getfd
-439	common	process_madvise			sys_process_madvise
+439	common	process_madvise			sys_process_madvise		compat_sys_process_madvise
diff --git a/arch/x86/entry/syscalls/syscall_32.tbl b/arch/x86/entry/syscalls/syscall_32.tbl
index 90950255ae5c..fd8032ddffb2 100644
--- a/arch/x86/entry/syscalls/syscall_32.tbl
+++ b/arch/x86/entry/syscalls/syscall_32.tbl
@@ -442,4 +442,4 @@
 435	i386	clone3			sys_clone3
 437	i386	openat2			sys_openat2
 438	i386	pidfd_getfd		sys_pidfd_getfd
-439	i386	process_madvise		sys_process_madvise
+439	i386	process_madvise		sys_process_madvise		compat_sys_process_madvise
diff --git a/arch/x86/entry/syscalls/syscall_64.tbl b/arch/x86/entry/syscalls/syscall_64.tbl
index bcf0d6d0dffe..3ff4be446922 100644
--- a/arch/x86/entry/syscalls/syscall_64.tbl
+++ b/arch/x86/entry/syscalls/syscall_64.tbl
@@ -359,7 +359,7 @@
 435	common	clone3			sys_clone3
 437	common	openat2			sys_openat2
 438	common	pidfd_getfd		sys_pidfd_getfd
-439	common	process_madvise		sys_process_madvise
+439	64	process_madvise		sys_process_madvise
 
 #
 # x32-specific system call numbers start at 512 to avoid cache impact
@@ -403,3 +403,4 @@
 545	x32	execveat		compat_sys_execveat
 546	x32	preadv2			compat_sys_preadv64v2
 547	x32	pwritev2		compat_sys_pwritev64v2
+548	x32	process_madvise		compat_sys_process_madvise
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 0480ba4db592..1134ba3e61d0 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -820,6 +820,9 @@ asmlinkage long compat_sys_pwritev64v2(unsigned long fd,
 		unsigned long vlen, loff_t pos, rwf_t flags);
 #endif
 
+asmlinkage ssize_t compat_sys_process_madvise(int which,
+		compat_pid_t upid, const struct compat_iovec __user *vec,
+		unsigned long vlen, int behavior, unsigned long flags);
 
 /*
  * Deprecated system calls which are still defined in
diff --git a/include/uapi/asm-generic/unistd.h b/include/uapi/asm-generic/unistd.h
index fa289b91410e..7fde54d6a8ac 100644
--- a/include/uapi/asm-generic/unistd.h
+++ b/include/uapi/asm-generic/unistd.h
@@ -856,7 +856,8 @@ __SYSCALL(__NR_openat2, sys_openat2)
 #define __NR_pidfd_getfd 438
 __SYSCALL(__NR_pidfd_getfd, sys_pidfd_getfd)
 #define __NR_process_madvise 439
-__SYSCALL(__NR_process_madvise, sys_process_madvise)
+__SC_COMP(__NR_process_madvise, sys_process_madvise, \
+		compat_sys_process_madvise)
 
 #undef __NR_syscalls
 #define __NR_syscalls 440
diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c
index 6c7332776e8e..16fca3a43633 100644
--- a/kernel/sys_ni.c
+++ b/kernel/sys_ni.c
@@ -281,6 +281,7 @@ COND_SYSCALL(munlockall);
 COND_SYSCALL(mincore);
 COND_SYSCALL(madvise);
 COND_SYSCALL(process_madvise);
+COND_SYSCALL_COMPAT(process_madvise);
 COND_SYSCALL(remap_file_pages);
 COND_SYSCALL(mbind);
 COND_SYSCALL_COMPAT(mbind);
diff --git a/mm/madvise.c b/mm/madvise.c
index 3082d7fa64ee..29bf2535624a 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -1195,7 +1195,7 @@ SYSCALL_DEFINE3(madvise, unsigned long, start, size_t, len_in, int, behavior)
 	return do_madvise(current, current->mm, start, len_in, behavior);
 }
 
-static int do_process_madvise(struct task_struct *target_task,
+static int process_madvise_vec(struct task_struct *target_task,
 		struct mm_struct *mm, struct iov_iter *iter, int behavior)
 {
 	struct iovec iovec;
@@ -1213,17 +1213,14 @@ static int do_process_madvise(struct task_struct *target_task,
 	return ret;
 }
 
-SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
-		const struct iovec __user *, vec, unsigned long, vlen,
-		int, behavior, unsigned long, flags)
+ssize_t do_process_madvise(int which, pid_t upid, struct iov_iter *iter,
+				int behavior, unsigned long flags)
 {
 	ssize_t ret;
 	struct pid *pid;
 	struct task_struct *task;
 	struct mm_struct *mm;
-	struct iovec iovstack[UIO_FASTIOV];
-	struct iovec *iov = iovstack;
-	struct iov_iter iter;
+	size_t total_len = iov_iter_count(iter);
 
 	if (flags != 0)
 		return -EINVAL;
@@ -1267,15 +1264,10 @@ SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
 		goto release_task;
 	}
 
-	ret = import_iovec(READ, vec, vlen, ARRAY_SIZE(iovstack), &iov, &iter);
-	if (ret >= 0) {
-		size_t total_len = iov_iter_count(&iter);
+	ret = process_madvise_vec(task, mm, iter, behavior);
+	if (ret >= 0)
+		ret = total_len - iov_iter_count(iter);
 
-		ret = do_process_madvise(task, mm, &iter, behavior);
-		if (ret >= 0)
-			ret = total_len - iov_iter_count(&iter);
-		kfree(iov);
-	}
 	mmput(mm);
 release_task:
 	put_task_struct(task);
@@ -1283,3 +1275,41 @@ SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
 	put_pid(pid);
 	return ret;
 }
+
+SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
+		const struct iovec __user *, vec, unsigned long, vlen,
+		int, behavior, unsigned long, flags)
+{
+	ssize_t ret;
+	struct iovec iovstack[UIO_FASTIOV];
+	struct iovec *iov = iovstack;
+	struct iov_iter iter;
+
+	ret = import_iovec(READ, vec, vlen, ARRAY_SIZE(iovstack), &iov, &iter);
+	if (ret >= 0) {
+		ret = do_process_madvise(which, upid, &iter, behavior, flags);
+		kfree(iov);
+	}
+	return ret;
+}
+
+#ifdef CONFIG_COMPAT
+COMPAT_SYSCALL_DEFINE6(process_madvise, int, which, compat_pid_t, upid,
+		const struct compat_iovec __user *, vec, unsigned long, vlen,
+		int, behavior, unsigned long, flags)
+
+{
+	ssize_t ret;
+	struct iovec iovstack[UIO_FASTIOV];
+	struct iovec *iov = iovstack;
+	struct iov_iter iter;
+
+	ret = compat_import_iovec(READ, vec, vlen, ARRAY_SIZE(iovstack),
+				&iov, &iter);
+	if (ret >= 0) {
+		ret = do_process_madvise(which, upid, &iter, behavior, flags);
+		kfree(iov);
+	}
+	return ret;
+}
+#endif
diff mbox series

Patch

diff --git a/mm/madvise.c b/mm/madvise.c
index 097506466fdc..3082d7fa64ee 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -1195,20 +1195,39 @@  SYSCALL_DEFINE3(madvise, unsigned long, start, size_t, len_in, int, behavior)
 	return do_madvise(current, current->mm, start, len_in, behavior);
 }
 
-SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid, unsigned long, start,
-		size_t, len_in, int, behavior, unsigned long, flags)
+static int do_process_madvise(struct task_struct *target_task,
+		struct mm_struct *mm, struct iov_iter *iter, int behavior)
 {
-	int ret;
+	struct iovec iovec;
+	int ret = 0;
+
+	while (iov_iter_count(iter)) {
+		iovec = iov_iter_iovec(iter);
+		ret = do_madvise(target_task, mm, (unsigned long)iovec.iov_base,
+					iovec.iov_len, behavior);
+		if (ret < 0)
+			break;
+		iov_iter_advance(iter, iovec.iov_len);
+	}
+
+	return ret;
+}
+
+SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid,
+		const struct iovec __user *, vec, unsigned long, vlen,
+		int, behavior, unsigned long, flags)
+{
+	ssize_t ret;
 	struct pid *pid;
 	struct task_struct *task;
 	struct mm_struct *mm;
+	struct iovec iovstack[UIO_FASTIOV];
+	struct iovec *iov = iovstack;
+	struct iov_iter iter;
 
 	if (flags != 0)
 		return -EINVAL;
 
-	if (!process_madvise_behavior_valid(behavior))
-		return -EINVAL;
-
 	switch (which) {
 	case P_PID:
 		if (upid <= 0)
@@ -1236,13 +1255,27 @@  SYSCALL_DEFINE6(process_madvise, int, which, pid_t, upid, unsigned long, start,
 		goto put_pid;
 	}
 
+	if (task->mm != current->mm &&
+			!process_madvise_behavior_valid(behavior)) {
+		ret = -EINVAL;
+		goto release_task;
+	}
+
 	mm = mm_access(task, PTRACE_MODE_ATTACH_FSCREDS);
 	if (IS_ERR_OR_NULL(mm)) {
 		ret = IS_ERR(mm) ? PTR_ERR(mm) : -ESRCH;
 		goto release_task;
 	}
 
-	ret = do_madvise(task, mm, start, len_in, behavior);
+	ret = import_iovec(READ, vec, vlen, ARRAY_SIZE(iovstack), &iov, &iter);
+	if (ret >= 0) {
+		size_t total_len = iov_iter_count(&iter);
+
+		ret = do_process_madvise(task, mm, &iter, behavior);
+		if (ret >= 0)
+			ret = total_len - iov_iter_count(&iter);
+		kfree(iov);
+	}
 	mmput(mm);
 release_task:
 	put_task_struct(task);