Skip to contentSkip to conversation

msgrcv: use freezable blocking call

DISCUSSION Maninder Singh 3 participants 3 messages Apr 1, 2015 High-resolution timers, timer wheel, clockevents

lore
Share & tools
lore

Maninder Singh

author Reviewed-by Ajeet Yadav Reviewed-by Ajeet Yadav Reviewed-by Ajeet Yadav
#0 lore raw

Hi Andrew,
Thanks for making new patch, Actually there is some problem with our mail editor. It changes tabs with spaces and corrupts the patch, we are solving the same at our end. Thats why i am sending you signed -off by only for both patches.

1. For msgrcv: use freezable blocking call

2. For restart_syscall: use freezable blocking call

----> For Peter's Review comment:- This is what, no why mentioned

This call was selected to be converted to a freezable call because it doesn't hold any locks or release any resources when interrupted that might be needed by another freezing task or a kernel driver during suspend, and is a common site where idle userspace tasks are blocked.

Thanks
Maninder Singh

Presumably Peter's review comments for "restart_syscall: use freezable blocking call" also apply here.

Please send your signed-off-by: for both patches, as detailed in Documentation/SubmittingPatches section 11, thanks.

From: Yogesh Gaur <[email protected]>
Subject: ipc/msg.c: use freezable blocking call

Avoid waking up every thread sleeping in a msgrcv call during suspend and resume by calling a freezable blocking call. Previous patches modified the freezer to avoid sending wakeups to threads that are blocked in freezable blocking calls.

Ref: https://lkml.org/lkml/2013/5/1/424

Backtrace: 
[<c03e3924>] (__schedule+0x0/0x5d8) from [<c03e3f88>] (schedule+0x8c/0x90)
[<c03e3efc>] (schedule+0x0/0x90) from [<c01ef9f8>] (do_msgrcv+0x2e0/0x368)
[<c01ef718>] (do_msgrcv+0x0/0x368) from [<c01efaac>] (SyS_msgrcv+0x2c/0x38)
[<c01efa80>] (SyS_msgrcv+0x0/0x38) from [<c001a180>] (ret_fast_syscall+0x0/0x48)

tPlay0Cb2 R running 0 297 204 0x00000001

---

 ipc/msg.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
ipc/msg.c renamed from ipc/msg.c~msgrcv-use-freezable-blocking-call +2 -1
diff -puN ipc/msg.c~msgrcv-use-freezable-blocking-call ipc/msg.c
@@ -37,6 +37,7 @@
 #include <linux/rwsem.h>
 #include <linux/nsproxy.h>
 #include <linux/ipc_namespace.h>
+#include <linux/freezer.h>
 
 #include <asm/current.h>
 #include <linux/uaccess.h>
@@ -915,7 +916,7 @@ long do_msgrcv(int msqid, void __user *b
 
 		ipc_unlock_object(&msq->q_perm);
 		rcu_read_unlock();
-		schedule();
+		freezable_schedule();
 
 		/* Lockless receive, part 1:
 		 * Disable preemption.  We don't hold a reference to the queue

_ ÿôèº{.nÇ+‰·Ÿ®‰­†+%ŠËÿ±éݶ¥Šwÿº{.nÇ+‰·¥Š{±þG«�éÿŠ{ayºʇڙë,j­¢f£¢·hš�ï�êÿ‘êçz_è®(­éšŽŠÝ¢j"�ú¶m§ÿÿ¾«þG«�éÿ¢¸?™¨è­Ú&£ø§~�á¶iO•æ¬z·švØ^¶m§ÿÿà ÿ¶ìÿ¢¸?–I¥

#1 lore raw
Quoted 4 lines · Maninder Singh

Hi Andrew,
Thanks for making new patch, Actually there is some problem with our mail editor. It changes tabs with spaces and corrupts the patch, we are solving the same at our end. Thats why i am sending you signed -off by only for both patches.

Sort it and resend, no real great hurry with these patches, right?

Quoted 4 lines

1. For msgrcv: use freezable blocking call

Did you really pass around that patch through 3 people or did it take all three of you to modify those two lines?

Should some of those SoBs be a reviewed-by perhaps?

Quoted 7 lines

----> For Peter's Review comment:- This is what, no why mentioned

This call was selected to be converted to a freezable call because it doesn't hold any locks or release any resources when interrupted that might be needed by another freezing task or a kernel driver during suspend, and is a common site where idle userspace tasks are blocked.

Please put such things in the Changelog so that we can see you've thought about things.

Reviewed-by Ajeet Yadav Reviewed-by Ajeet Yadav
#2 lore raw
Quoted 7 lines · On Wed, 1 Apr 2015 09:35:55 +0200 Peter Zijlstra <[email protected]> wrote:
Quoted 4 lines · Maninder Singh

Hi Andrew,
Thanks for making new patch, Actually there is some problem with our mail editor. It changes tabs with spaces and corrupts the patch, we are solving the same at our end. Thats why i am sending you signed -off by only for both patches.

Sort it and resend, no real great hurry with these patches, right?

I tend not to bother too much about occasional messy patches. These ones appear to be be the first patches from these contributors and fixing them up only takes a couple of minutes. If Maninder's team expects to send more patches in the future then yes, please fix this stuff. But for now, the important thing is to get these kernel problems sorted out.

Quoted 13 lines
Quoted 4 lines

1. For msgrcv: use freezable blocking call

Did you really pass around that patch through 3 people or did it take all three of you to modify those two lines?

Should some of those SoBs be a reviewed-by perhaps?

Quoted 2 lines

----> For Peter's Review comment:- This is what, no why mentioned

Quoted 8 lines
Quoted 5 lines

This call was selected to be converted to a freezable call because it doesn't hold any locks or release any resources when interrupted that might be needed by another freezing task or a kernel driver during suspend, and is a common site where idle userspace tasks are blocked.

Please put such things in the Changelog so that we can see you've thought about things.

I have made that change.

Maninder, we currently have yourself as the primary author of "restart_syscall: use freezable blocking call". Is that correct, or should that be Yogesh Gaur?

Below are my latest copies of these two patches. How do they look?

From: Yogesh Gaur <[email protected]>
Subject: ipc/msg.c: use freezable blocking call

Avoid waking up every thread sleeping in a msgrcv call during suspend and resume by calling a freezable blocking call. Previous patches modified the freezer to avoid sending wakeups to threads that are blocked in freezable blocking calls.

Ref: https://lkml.org/lkml/2013/5/1/424

Backtrace: 
[<c03e3924>] (__schedule+0x0/0x5d8) from [<c03e3f88>] (schedule+0x8c/0x90)
[<c03e3efc>] (schedule+0x0/0x90) from [<c01ef9f8>] (do_msgrcv+0x2e0/0x368)
[<c01ef718>] (do_msgrcv+0x0/0x368) from [<c01efaac>] (SyS_msgrcv+0x2c/0x38)
[<c01efa80>] (SyS_msgrcv+0x0/0x38) from [<c001a180>] (ret_fast_syscall+0x0/0x48)

tPlay0Cb2 R running 0 297 204 0x00000001

This call was selected to be converted to a freezable call because it doesn't hold any locks or release any resources when interrupted that might be needed by another freezing task or a kernel driver during suspend, and is a common site where idle userspace tasks are blocked.

---

 ipc/msg.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
ipc/msg.c renamed from ipc/msg.c~msgrcv-use-freezable-blocking-call +2 -1
diff -puN ipc/msg.c~msgrcv-use-freezable-blocking-call ipc/msg.c
@@ -37,6 +37,7 @@
 #include <linux/rwsem.h>
 #include <linux/nsproxy.h>
 #include <linux/ipc_namespace.h>
+#include <linux/freezer.h>
 
 #include <asm/current.h>
 #include <linux/uaccess.h>
@@ -915,7 +916,7 @@ long do_msgrcv(int msqid, void __user *b
 
 		ipc_unlock_object(&msq->q_perm);
 		rcu_read_unlock();
-		schedule();
+		freezable_schedule();
 
 		/* Lockless receive, part 1:
 		 * Disable preemption.  We don't hold a reference to the queue

_

From: Maninder Singh <[email protected]>
Subject: kernel/time/hrtimer.c: restart_syscall: use freezable blocking call

Avoid waking up every thread sleeping in a restart_syscall call during suspend and resume by calling a freezable blocking call. Previous patches modified the freezer to avoid sending wakeups to threads that are blocked in freezable blocking calls.

Ref: https://lkml.org/lkml/2013/5/1/424

Backtrace: 
[<c03e3924>] (__schedule+0x0/0x5d8) from [<c03e3f88>] (schedule+0x8c/0x90)
[<c03e3efc>] (schedule+0x0/0x90) from [<c03e3150>] (schedule_hrtimeout_range_clock+0xdc/0x110)
[<c03e3074>] (schedule_hrtimeout_range_clock+0x0/0x110) from [<c03e31a0>] (schedule_hrtimeout_range+0x1c/0x20)
 r9:d16c9be0 r8:8b7d9c2c r7:00000000 r6:00000000 r5:d16c8028
[<c03e3184>] (schedule_hrtimeout_range+0x0/0x20) from [<c015778c>] (poll_schedule_timeout+0x48/0x6c)
[<c0157744>] (poll_schedule_timeout+0x0/0x6c) from [<c0158994>] (do_sys_poll+0x2c8/0x378) r5:d16c9f78 r4:00000000
[<c01586cc>] (do_sys_poll+0x0/0x378) from [<c0158a84>] (do_restart_poll+0x40/0x5c)
[<c0158a44>] (do_restart_poll+0x0/0x5c) from [<c005710c>] (sys_restart_syscall+0x2c/0x30) r4:fffffe7a
[<c00570e0>] (sys_restart_syscall+0x0/0x30) from [<c001a180>] (ret_fast_syscall+0x0/0x48)

This call was selected to be converted to a freezable call because it doesn't hold any locks or release any resources when interrupted that might be needed by another freezing task or a kernel driver during suspend, and is a common site where idle userspace tasks are blocked.

---

 kernel/time/hrtimer.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/time/hrtimer.c renamed from kernel/time/hrtimer.c~restart_syscall-use-freezable-blocking-call +2 -2
diff -puN kernel/time/hrtimer.c~restart_syscall-use-freezable-blocking-call kernel/time/hrtimer.c
@@ -1767,7 +1767,7 @@ schedule_hrtimeout_range_clock(ktime_t *
 	 * A NULL parameter means "infinite"
 	 */
 	if (!expires) {
-		schedule();
+		freezable_schedule();
 		return -EINTR;
 	}
 
@@ -1781,7 +1781,7 @@ schedule_hrtimeout_range_clock(ktime_t *
 		t.task = NULL;
 
 	if (likely(t.task))
-		schedule();
+		freezable_schedule();
 
 	hrtimer_cancel(&t.timer);
 	destroy_hrtimer_on_stack(&t.timer);

_