Skip to contentSkip to conversation

mm,hwpoison: Differentiate Action Required error in current and non current processes

PATCH Superseded Aili Yao 2 participants 3 messages Jan 20, 2021

Subsystems (2)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Aili Yao

author
#0 lore raw

PATCH v3 1/1 mm,hwpoison: Differentiate Action Required error in current and non current processes

When a memory uncorrected error is triggered by process A who accessed the address with error; It's Action Required Case for only current process which triggered this.this Action Required case means Action optional to other process who share the same page. Usually, kill current process will be sufficient, other process sharing the same page will get be signaled when they really touch the poisoned page.

But there is another scenario that other processes
sharing the same page want to be signaled early with PF_MCE_EARLY set, In this case, we should get them into kill list and signal BUS_MCEERR_AO to them.

So in this patch, task_early_kill will check current process if force_early is set, and if not current,check find_early_kill_thread to see if there is PF_MCE_EARLY process which cares the error.

In kill_proc, BUS_MCEERR_AR is only send to current, other process in kill list will be signaled BUS_MCEERR_AO.

---
 mm/memory-failure.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
mm/memory-failure.c +8 -4
index 5a38e9eade94..2d6047780466 100644
@@ -243,9 +243,12 @@ static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags)
 			pfn, t->comm, t->pid);
 
 	if (flags & MF_ACTION_REQUIRED) {
-		WARN_ON_ONCE(t != current);
-		ret = force_sig_mceerr(BUS_MCEERR_AR,
+		if (tk->tsk == current)
+			ret = force_sig_mceerr(BUS_MCEERR_AR,
 					 (void __user *)tk->addr, addr_lsb);
+		else
+			ret = send_sig_mceerr(BUS_MCEERR_AO, (void __user *)tk->addr,
+				addr_lsb, t);
 	} else {
 		/*
 		 * Don't use force here, it's convenient if the signal
@@ -454,11 +457,12 @@ static struct task_struct *task_early_kill(struct task_struct *tsk,
 		/*
 		 * Comparing ->mm here because current task might represent
 		 * a subthread, while tsk always points to the main thread.
+		 * If tsk is not current, we need to fallback to
+		 * find_early_kill_thread checking whether other processes with
+		 * PF_MCE_EARLY set still care the error.
 		 */
 		if (tsk->mm == current->mm)
 			return current;
-		else
-			return NULL;
 	}
 	return find_early_kill_thread(tsk);
 }
Signature · 1 line
2.25.1
#1 lore raw
Quoted 16 lines · Aili Yao

When a memory uncorrected error is triggered by process A who accessed the address with error; It's Action Required Case for only current process which triggered this.this Action Required case means Action optional to other process who share the same page. Usually, kill current process will be sufficient, other process sharing the same page will get be signaled when they really touch the poisoned page.

But there is another scenario that other processes
sharing the same page want to be signaled early with PF_MCE_EARLY set, In this case, we should get them into kill list and signal BUS_MCEERR_AO to them.

So in this patch, task_early_kill will check current process if force_early is set, and if not current,check find_early_kill_thread to see if there is PF_MCE_EARLY process which cares the error.

Quoted 5 lines

In kill_proc, BUS_MCEERR_AR is only send to current, other process in kill list will be signaled BUS_MCEERR_AO.

"Reviewed-by" tag has some special meaning in Linux kernel development, and that should be tagged explicitly by reviewers. So please wait a little more to add this :)

See the document for more details:
https://www.kernel.org/doc/html/v5.10/process/submitting-patches.html#using-reported-by-tested-by-reviewed-by-suggested-by-and-fixes

Quoted 25 lines
---
 mm/memory-failure.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
mm/memory-failure.c +5 -2
index 5a38e9eade94..2d6047780466 100644
@@ -243,9 +243,12 @@ static int kill_proc(struct to_kill *tk, unsigned long pfn, int flags)
 			pfn, t->comm, t->pid);
 
 	if (flags & MF_ACTION_REQUIRED) {
-		WARN_ON_ONCE(t != current);
-		ret = force_sig_mceerr(BUS_MCEERR_AR,
+		if (tk->tsk == current)
+			ret = force_sig_mceerr(BUS_MCEERR_AR,
 					 (void __user *)tk->addr, addr_lsb);
+		else
+			ret = send_sig_mceerr(BUS_MCEERR_AO, (void __user *)tk->addr,
+				addr_lsb, t);
 	} else {
 		/*
 		 * Don't use force here, it's convenient if the signal

mm/memory-failure.c static struct task_struct *task_early_kill(struct task_struct *tsk, · line 463

Quoted 7 lines
mm/memory-failure.c +3 -0
@@ -454,11 +457,12 @@ static struct task_struct *task_early_kill(struct task_struct *tsk,
 		/*
 		 * Comparing ->mm here because current task might represent
 		 * a subthread, while tsk always points to the main thread.
+		 * If tsk is not current, we need to fallback to
+		 * find_early_kill_thread checking whether other processes with
+		 * PF_MCE_EARLY set still care the error.
 		 */

Sorry for my unclearness, I meant to ask for updating the comment below (especially "Note that ..." part), which gets obsolete by your patch.

Quoted 9 lines
mm/memory-failure.c +0 -0
 /*
 * Determine whether a given process is "early kill" process which expects
 * to be signaled when some page under the process is hwpoisoned.
 * Return task_struct of the dedicated thread (main thread unless explicitly
 * specified) if the process is "early kill," and otherwise returns NULL.
 *
 * Note that the above is true for Action Optional case, but not for Action
 * Required case where SIGBUS should sent only to the current thread.
 */
Quoted 3 lines

static struct task_struct *task_early_kill(struct task_struct *tsk,
int force_early)
{

Thanks,
Naoya Horiguchi

Quoted 10 lines
mm/memory-failure.c +0 -2
 		if (tsk->mm == current->mm)
 			return current;
-		else
-			return NULL;
 	}
 	return find_early_kill_thread(tsk);
 }
Signature · 1 line
2.25.1
author
#2 lore raw
Quoted 8 lines · On Wed, 20 Jan 2021 06:37:04 +0000 HORIGUCHI NAOYA(堀口 直也)…

"Reviewed-by" tag has some special meaning in Linux kernel development, and that should be tagged explicitly by reviewers. So please wait a little more to add this :)

See the document for more details:
https://www.kernel.org/doc/html/v5.10/process/submitting-patches.html#using-reported-by-tested-by-reviewed-by-suggested-by-and-fixes

got it, Thanks

Quoted 3 lines

Sorry for my unclearness, I meant to ask for updating the comment below (especially "Note that ..." part), which gets obsolete by your patch.

I have submitted a v4 patch,

Thank you!

Signature · 3 lines
Best Regards!

Aili Yao