Skip to contentSkip to conversation

Two x86 fixes

PATCH Superseded Ammar Faizi 2 participants 13 messages Mar 10, 2022 – Mar 28, 2022

Subsystems (4)

v5 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

v5 cover Two x86 fixes

Hi,

Two x86 fixes in this series.

1) x86/delay: Fix the wrong Assembly constraint in delay_loop() function.
2) x86/MCE/AMD: Fix memory leak when threshold_create_bank() fails.

## Changelog

v5:
- Mark patch #1 for stable.
- Commit message improvement for patch #1 and #2.
- Fold in changes from Yazen and Alviro (for patch #2).

v4:
- Address comment from Greg, sha1 commit Fixes only needs to be 12 chars.
- Add the author of the fixed commit to the CC list.

v3:
- Fold in changes from Alviro, the previous version is still leaking @bank[n].

v2:
- Fix wrong copy/paste.

---

Ammar Faizi (2):
x86/delay: Fix the wrong asm constraint in delay_loop() x86/MCE/AMD: Fix memory leak when threshold_create_bank() fails

 arch/x86/kernel/cpu/mce/amd.c | 32 +++++++++++++++++++-------------
 arch/x86/lib/delay.c          |  4 ++--
 2 files changed, 21 insertions(+), 15 deletions(-)

base-commit: 7e57714cd0ad2d5bb90e50b5096a0e671dec1ef3

Signature · 1 line
Ammar Faizi
#1 lore raw

PATCH v5 1/2 x86/delay: Fix the wrong asm constraint in `delay_loop()`

The asm constraint does not reflect that the asm statement can modify the value of @loops. But the asm statement in delay_loop() does modify the @loops.

Specifiying the wrong constraint may lead to undefined behavior, it may clobber random stuff (e.g. local variable, important temporary value in regs, etc.).

Fix this by changing the constraint from "a" (as an input) to "+a" (as an input and output).

Cc David Laight <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Dave Hansen <[email protected]>
Cc "H. Peter Anvin" <[email protected]>
Cc Jiri Hladky <[email protected]>
Cc [email protected] # v2.6.27+
Fixes e01b70ef3eb ("x86: fix bug in arch/i386/lib/delay.c file, delay_loop function")
---
 arch/x86/lib/delay.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/lib/delay.c +2 -2
index 65d15df6212d..0e65d00e2339 100644
@@ -54,8 +54,8 @@ static void delay_loop(u64 __loops)
 		"	jnz 2b		\n"
 		"3:	dec %0		\n"
 
-		: /* we don't need output */
-		:"a" (loops)
+		: "+a" (loops)
+		:
 	);
 }
 
Signature · 1 line
2.32.0
maintainer
#2 lore raw

Re: [PATCH v5 1/2] x86/delay: Fix the wrong asm constraint in `delay_loop()`

Quoted 7 lines · Ammar Faizi

The asm constraint does not reflect that the asm statement can modify the value of @loops. But the asm statement in delay_loop() does modify the @loops.

Specifiying the wrong constraint may lead to undefined behavior, it may clobber random stuff (e.g. local variable, important temporary value in regs, etc.).

This is especially dangerous when the compiler decides to inline the function and since it doesn't know that the value gets modified, it might decide to use it from a register directly without reloading it.

Add that to the commit message pls.

Quoted 10 lines

Fix this by changing the constraint from "a" (as an input) to "+a" (as an input and output).

Cc David Laight <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Dave Hansen <[email protected]>
Cc "H. Peter Anvin" <[email protected]>
Cc Jiri Hladky <[email protected]>

All those Ccs in the commit message are not really needed - get_maintainers.pl gives the correct list already.

Quoted 1 line
Cc [email protected] # v2.6.27+

I don't see the need for the stable Cc. Or do you have a case where a corruption really does happen?

Quoted 1 line
Fixes e01b70ef3eb ("x86: fix bug in arch/i386/lib/delay.c file, delay_loop function")

Commit sha1 (e01b70ef3eb) needs to be at least 12 chars long:

e01b70ef3eb3 ("x86: fix bug in arch/i386/lib/delay.c file, delay_loop function")

This is best fixed by doing:

[core]

        abbrev = 12

in your .git/config

Quoted 10 lines
---
 arch/x86/lib/delay.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/lib/delay.c +0 -0
index 65d15df6212d..0e65d00e2339 100644

arch/x86/lib/delay.c static void delay_loop(u64 __loops) · line 59

Quoted 8 lines
arch/x86/lib/delay.c +2 -2
@@ -54,8 +54,8 @@ static void delay_loop(u64 __loops)
 		"	jnz 2b		\n"
 		"3:	dec %0		\n"
 
-		: /* we don't need output */
-		:"a" (loops)
+		: "+a" (loops)
+		:
 	);

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#3 lore raw

Re: [PATCH v5 1/2] x86/delay: Fix the wrong asm constraint in `delay_loop()`

Quoted 12 lines · Borislav Petkov

All those Ccs in the commit message are not really needed - get_maintainers.pl gives the correct list already.

Quoted 1 line
Cc [email protected] # v2.6.27+

I don't see the need for the stable Cc. Or do you have a case where a corruption really does happen?

Quoted 1 line
Fixes e01b70ef3eb ("x86: fix bug in arch/i386/lib/delay.c file, delay_loop function")

Commit sha1 (e01b70ef3eb) needs to be at least 12 chars long:

Quoted 8 lines

e01b70ef3eb3 ("x86: fix bug in arch/i386/lib/delay.c file, delay_loop function")

This is best fixed by doing:

[core]

         abbrev = 12

in your .git/config

Will fix that in the v6.

Signature · 1 line
Ammar Faizi
#4 lore raw

Re: [PATCH v5 1/2] x86/delay: Fix the wrong asm constraint in `delay_loop()`

In reply to Borislav Petkov
Quoted 9 lines · Borislav Petkov
Quoted 7 lines · Ammar Faizi

The asm constraint does not reflect that the asm statement can modify the value of @loops. But the asm statement in delay_loop() does modify the @loops.

Specifiying the wrong constraint may lead to undefined behavior, it may clobber random stuff (e.g. local variable, important temporary value in regs, etc.).

Quoted 5 lines

This is especially dangerous when the compiler decides to inline the function and since it doesn't know that the value gets modified, it might decide to use it from a register directly without reloading it.

Add that to the commit message pls.

Will add that in the v6.

Quoted 4 lines
Quoted 1 line
Cc [email protected] # v2.6.27+

I don't see the need for the stable Cc. Or do you have a case where a corruption really does happen?

I don't find any visible issue on this. But that's undefined behavior, different compiler may yield different result (e.g. there is no guarantee newer compilers will produce the appropriate result due to UB). So it's not something we should rely on.

============
Side note for inline:
Even if it is not inlined, it's still dangerous, because if the compiler is able to see that the function to be called doesn't clobber some call-clobbered regs, the compiler can assume the call-clobbered regs are not clobbered and it reuses the value without reloading.

See the example from Alviro here:

https://lore.kernel.org/lkml/CAOG64qPgTv5tQNknuG9d-=oL2EPQQ1ys7xu2FoBpNLyzv1qYzA@mail.gmail.com/

Signature · 1 line
Ammar Faizi
#6 lore raw

PATCH v5 2/2 x86/MCE/AMD: Fix memory leak when `threshold_create_bank()` fails

In reply to Ammar Faizi

In mce_threshold_create_device(), if threshold_create_bank() fails, the @bp will be leaked, because the call to mce_threshold_remove_device() will not free the @bp. mce_threshold_remove_device() frees @threshold_banks. At that point, the @bp has not been written to @threshold_banks, @threshold_banks is NULL, so the call is just a nop.

Fix this by extracting the cleanup part into a new static function _mce_threshold_remove_device(), then call it from create/remove device functions.

Also, eliminate the "goto out_err", just early return inside the loop if the creation fails.

Cc Borislav Petkov <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Fixes 6458de97fc15 ("x86/mce/amd: Straighten CPU hotplug path")
---
 arch/x86/kernel/cpu/mce/amd.c | 32 +++++++++++++++++++-------------
 1 file changed, 19 insertions(+), 13 deletions(-)
arch/x86/kernel/cpu/mce/amd.c +19 -13
index 9f4b508886dd..e492efab68a2 100644
@@ -1293,10 +1293,23 @@ static void threshold_remove_bank(struct threshold_bank *bank)
 	kfree(bank);
 }
 
+static void _mce_threshold_remove_device(struct threshold_bank **bp,
+					 unsigned int numbanks)
+{
+	unsigned int bank;
+
+	for (bank = 0; bank < numbanks; bank++) {
+		if (bp[bank]) {
+			threshold_remove_bank(bp[bank]);
+			bp[bank] = NULL;
+		}
+	}
+	kfree(bp);
+}
+
 int mce_threshold_remove_device(unsigned int cpu)
 {
 	struct threshold_bank **bp = this_cpu_read(threshold_banks);
-	unsigned int bank, numbanks = this_cpu_read(mce_num_banks);
 
 	if (!bp)
 		return 0;
@@ -1307,13 +1320,7 @@ int mce_threshold_remove_device(unsigned int cpu)
 	 */
 	this_cpu_write(threshold_banks, NULL);
 
-	for (bank = 0; bank < numbanks; bank++) {
-		if (bp[bank]) {
-			threshold_remove_bank(bp[bank]);
-			bp[bank] = NULL;
-		}
-	}
-	kfree(bp);
+	_mce_threshold_remove_device(bp, this_cpu_read(mce_num_banks));
 	return 0;
 }
 
@@ -1350,15 +1357,14 @@ int mce_threshold_create_device(unsigned int cpu)
 		if (!(this_cpu_read(bank_map) & (1 << bank)))
 			continue;
 		err = threshold_create_bank(bp, cpu, bank);
-		if (err)
-			goto out_err;
+		if (err) {
+			_mce_threshold_remove_device(bp, numbanks);
+			return err;
+		}
 	}
 	this_cpu_write(threshold_banks, bp);
 
 	if (thresholding_irq_en)
 		mce_threshold_vector = amd_threshold_interrupt;
 	return 0;
-out_err:
-	mce_threshold_remove_device(cpu);
-	return err;
 }
Signature · 1 line
2.32.0
maintainer
#7 lore raw

Re: [PATCH v5 2/2] x86/MCE/AMD: Fix memory leak when `threshold_create_bank()` fails

Quoted 10 lines · Ammar Faizi

In mce_threshold_create_device(), if threshold_create_bank() fails, the @bp will be leaked, because the call to mce_threshold_remove_device() will not free the @bp. mce_threshold_remove_device() frees @threshold_banks. At that point, the @bp has not been written to @threshold_banks, @threshold_banks is NULL, so the call is just a nop.

Fix this by extracting the cleanup part into a new static function _mce_threshold_remove_device(), then call it from create/remove device functions.

Quoted 8 lines

Also, eliminate the "goto out_err", just early return inside the loop if the creation fails.

Cc Borislav Petkov <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Fixes 6458de97fc15 ("x86/mce/amd: Straighten CPU hotplug path")

How did you decide this is the commit that this is fixing?

Quoted 1 line

That Link tag is not needed.

Quoted 3 lines

There's no "Co-authored-by".

The correct tag is described in

Documentation/process/submitting-patches.rst

Please make sure you've read that file before sending patches.

Quoted 5 lines
---
 arch/x86/kernel/cpu/mce/amd.c | 32 +++++++++++++++++++-------------
 1 file changed, 19 insertions(+), 13 deletions(-)

...

arch/x86/kernel/cpu/mce/amd.c int mce_threshold_create_device(unsigned int cpu) · line 1365

Quoted 12 lines
arch/x86/kernel/cpu/mce/amd.c +4 -2
@@ -1350,15 +1357,14 @@ int mce_threshold_create_device(unsigned int cpu)
 		if (!(this_cpu_read(bank_map) & (1 << bank)))
 			continue;
 		err = threshold_create_bank(bp, cpu, bank);
-		if (err)
-			goto out_err;
+		if (err) {
+			_mce_threshold_remove_device(bp, numbanks);
+			return err;
+		}
 	}
 	this_cpu_write(threshold_banks, bp);

Do I see it correctly that the publishing of the @bp pointer - i.e., this line - should be moved right above the for loop?

Then mce_threshold_remove_device() would properly free it in the error case and your patch turns into a oneliner?

And then your Fixes: tag would be correct too...

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#8 lore raw

Re: [PATCH v5 2/2] x86/MCE/AMD: Fix memory leak when `threshold_create_bank()` fails

On 3/28/22 5:52 AM, Borislav Petkov wrote:
[...]

Quoted 3 lines
Quoted 1 line
Fixes 6458de97fc15 ("x86/mce/amd: Straighten CPU hotplug path")

How did you decide this is the commit that this is fixing?

I examined the history in those lines by git blame. Will recheck after the below doubt is cleared.

Quoted 8 lines
Quoted 1 line

That Link tag is not needed.

Quoted 3 lines
Quoted 5 lines

There's no "Co-authored-by".

The correct tag is described in

Documentation/process/submitting-patches.rst

Will fix them in the v6.

Quoted 18 lines

...

arch/x86/kernel/cpu/mce/amd.c int mce_threshold_create_device(unsigned int cpu) · line 1365

Quoted 12 lines
arch/x86/kernel/cpu/mce/amd.c +4 -2
@@ -1350,15 +1357,14 @@ int mce_threshold_create_device(unsigned int cpu)
  		if (!(this_cpu_read(bank_map) & (1 << bank)))
  			continue;
  		err = threshold_create_bank(bp, cpu, bank);
-		if (err)
-			goto out_err;
+		if (err) {
+			_mce_threshold_remove_device(bp, numbanks);
+			return err;
+		}
  	}
  	this_cpu_write(threshold_banks, bp);

Do I see it correctly that the publishing of the @bp pointer - i.e., this line - should be moved right above the for loop?

Quoted 2 lines

Then mce_threshold_remove_device() would properly free it in the error case and your patch turns into a oneliner?

Previously, in v4 I did that too. But after discussion with Yazen, we got a conclusion that placing this_cpu_write(threshold_banks, bp); before the for loop is not the right thing to do.

Quoted 1 line

And then your Fixes: tag would be correct too...

The reason is based on the discussion with Yazen, the full discussion can be read in the Link tag above.

==================
The point is:

Quoted 4 lines · Yazen Ghannam

The threshold interrupt handler uses this pointer. I think the goal here is to set this pointer when the list is fully formed and clear this pointer before making any changes to the list. Otherwise, the interrupt handler will operate on incomplete data if an interrupt comes in the middle of these updates.

==================

Also, looking at the comment in mce_threshold_remove_device() function:

	/*
	 * Clear the pointer before cleaning up, so that the interrupt won't
	 * touch anything of this.
	 */
	this_cpu_write(threshold_banks, NULL);

I think it's reasonable to place this_cpu_write(threshold_banks, bp); after the "for loop" on the creation process for the similar reason. In short, don't let the interrupt sees incomplete data.

Although, I am not sure if that 100% guarantees mce_threshold_remove_device() will not mess up with the interrupt (e.g. freeing the data while the interrupt reading it), unless we're using RCU stuff.

What do you think?

Signature · 1 line
Ammar Faizi
maintainer
#9 lore raw

Re: [PATCH v5 2/2] x86/MCE/AMD: Fix memory leak when `threshold_create_bank()` fails

Quoted 5 lines · Ammar Faizi

Although, I am not sure if that 100% guarantees mce_threshold_remove_device() will not mess up with the interrupt (e.g. freeing the data while the interrupt reading it), unless we're using RCU stuff.

What do you think?

I would've said it doesn't matter but that thresholding device creation is part of hotplug and it can happen multiple times even *after* the interrupt vector has been set during setup so a potential teardown and concurrent thresholding interrupt firing might really hit in a not fully initialized/cleaned up state so yeah, let's do Yazen's thing.

The alternative would be the temporarily re-assign mce_threshold_vector to default_threshold_interrupt while setup is being done but that's not really necessary atm.

But call that helper function __threshold_remove_device().

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
maintainer
#11 lore raw
Quoted 8 lines · Ammar Faizi
Quoted 4 lines · Ammar Faizi

Two x86 fixes in this series.

1) x86/delay: Fix the wrong Assembly constraint in delay_loop() function.
2) x86/MCE/AMD: Fix memory leak when threshold_create_bank() fails.

Ping (1)!
Borislav? Thomas?

Yes, what's up?

Are those urgent fixes which break some use case or can you simply sit patiently and wait?

Because we have an upcoming merge window and we need to prepare for that. And there are real bugs that need fixing too.

So what's the rush here?

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#12 lore raw
Quoted 4 lines · Borislav Petkov

Yes, what's up?

Are those urgent fixes which break some use case or can you simply sit patiently and wait?

Sorry for pinging at the wrong time. Excuse my weekly ping. They are not urgent fixes. So no rush.

Quoted 2 lines

Because we have an upcoming merge window and we need to prepare for that. And there are real bugs that need fixing too.

Hopefully, the 5.17 release and 5.18 merge window go well.

Signature · 1 line
Ammar Faizi