[PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Shihao Ren]

In the funcs allocation failure path, the cleanup loop iterates with
'j' but frees objs[i].funcs, which uses the outer loop index 'i'
instead of the loop variable 'j'. As a result the just-failed (NULL)
entry is freed repeatedly, while the funcs buffers already allocated
for objs[0..i-1] are leaked.

Use objs[j].funcs so the previously allocated entries are correctly
released.

Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
Signed-off-by: Shihao Ren <rsh15355756202@163.com>
---
 scripts/livepatch/init.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/livepatch/init.c b/scripts/livepatch/init.c
index f14d8c8fb..16aff8f73 100644
--- a/scripts/livepatch/init.c
+++ b/scripts/livepatch/init.c
@@ -51,7 +51,7 @@ static int __init livepatch_mod_init(void)
 		if (!funcs) {
 			ret = -ENOMEM;
 			for (int j = 0; j < i; j++)
-				kfree(objs[i].funcs);
+				kfree(objs[j].funcs);
 			goto err_free_objs;
 		}
 
-- 
2.39.5

Re: [PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Miroslav Benes]

On Wed, 1 Jul 2026, Shihao Ren wrote:

> In the funcs allocation failure path, the cleanup loop iterates with
> 'j' but frees objs[i].funcs, which uses the outer loop index 'i'
> instead of the loop variable 'j'. As a result the just-failed (NULL)
> entry is freed repeatedly, while the funcs buffers already allocated
> for objs[0..i-1] are leaked.
> 
> Use objs[j].funcs so the previously allocated entries are correctly
> released.
> 
> Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
> Signed-off-by: Shihao Ren <rsh15355756202@163.com>

Acked-by: Miroslav Benes <mbenes@suse.cz>

M

Re: [PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Petr Mladek]

On Wed 2026-07-01 14:06:55, Shihao Ren wrote:
> In the funcs allocation failure path, the cleanup loop iterates with
> 'j' but frees objs[i].funcs, which uses the outer loop index 'i'
> instead of the loop variable 'j'. As a result the just-failed (NULL)
> entry is freed repeatedly, while the funcs buffers already allocated
> for objs[0..i-1] are leaked.
> 
> Use objs[j].funcs so the previously allocated entries are correctly
> released.
> 
> Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
> Signed-off-by: Shihao Ren <rsh15355756202@163.com>

Great catch!

Reviewed-by: Petr Mladek <pmladek@suse.com>

I assume that Josh would take this together with other klp-build
related changes.

Best Regards,
Petr

Re: [PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Miroslav Benes]

On Thu, 2 Jul 2026, Petr Mladek wrote:

> On Wed 2026-07-01 14:06:55, Shihao Ren wrote:
> > In the funcs allocation failure path, the cleanup loop iterates with
> > 'j' but frees objs[i].funcs, which uses the outer loop index 'i'
> > instead of the loop variable 'j'. As a result the just-failed (NULL)
> > entry is freed repeatedly, while the funcs buffers already allocated
> > for objs[0..i-1] are leaked.
> > 
> > Use objs[j].funcs so the previously allocated entries are correctly
> > released.
> > 
> > Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
> > Signed-off-by: Shihao Ren <rsh15355756202@163.com>
> 
> Great catch!
> 
> Reviewed-by: Petr Mladek <pmladek@suse.com>
> 
> I assume that Josh would take this together with other klp-build
> related changes.

I do not see it merged yet. Josh, could you take it, please?

Miroslav

Re: [PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Josh Poimboeuf]

On Mon, Aug 24, 2026 at 03:42:46PM +0200, Miroslav Benes wrote:
> On Thu, 2 Jul 2026, Petr Mladek wrote:
> 
> > On Wed 2026-07-01 14:06:55, Shihao Ren wrote:
> > > In the funcs allocation failure path, the cleanup loop iterates with
> > > 'j' but frees objs[i].funcs, which uses the outer loop index 'i'
> > > instead of the loop variable 'j'. As a result the just-failed (NULL)
> > > entry is freed repeatedly, while the funcs buffers already allocated
> > > for objs[0..i-1] are leaked.
> > > 
> > > Use objs[j].funcs so the previously allocated entries are correctly
> > > released.
> > > 
> > > Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
> > > Signed-off-by: Shihao Ren <rsh15355756202@163.com>
> > 
> > Great catch!
> > 
> > Reviewed-by: Petr Mladek <pmladek@suse.com>
> > 
> > I assume that Josh would take this together with other klp-build
> > related changes.
> 
> I do not see it merged yet. Josh, could you take it, please?

Sorry, I missed seeing this one, but I do have an identical patch
already queued:

  https://lore.kernel.org/aocdq0hr_iQTimG0@jpoimboe

-- 
Josh

Re: [PATCH] livepatch/klp-build: Fix wrong index in error cleanup loop

[Miroslav Benes]

On Thu, 27 Aug 2026, Josh Poimboeuf wrote:

> On Mon, Aug 24, 2026 at 03:42:46PM +0200, Miroslav Benes wrote:
> > On Thu, 2 Jul 2026, Petr Mladek wrote:
> > 
> > > On Wed 2026-07-01 14:06:55, Shihao Ren wrote:
> > > > In the funcs allocation failure path, the cleanup loop iterates with
> > > > 'j' but frees objs[i].funcs, which uses the outer loop index 'i'
> > > > instead of the loop variable 'j'. As a result the just-failed (NULL)
> > > > entry is freed repeatedly, while the funcs buffers already allocated
> > > > for objs[0..i-1] are leaked.
> > > > 
> > > > Use objs[j].funcs so the previously allocated entries are correctly
> > > > released.
> > > > 
> > > > Fixes: 59adee07b568 ("livepatch/klp-build: Add stub init code for livepatch modules")
> > > > Signed-off-by: Shihao Ren <rsh15355756202@163.com>
> > > 
> > > Great catch!
> > > 
> > > Reviewed-by: Petr Mladek <pmladek@suse.com>
> > > 
> > > I assume that Josh would take this together with other klp-build
> > > related changes.
> > 
> > I do not see it merged yet. Josh, could you take it, please?
> 
> Sorry, I missed seeing this one, but I do have an identical patch
> already queued:
> 
>   https://lore.kernel.org/aocdq0hr_iQTimG0@jpoimboe

Ok, thanks. I did not get that far in my INBOX after holiday yet :/

Miroslav