Skip to contentSkip to conversation

staging: r8188eu: Use kzalloc() with GFP_ATOMIC in atomic context

PATCH Superseded Fabio M. De Francesco 4 participants 7 messages Nov 1, 2021 – Nov 2, 2021 Staging

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 staging: r8188eu: Use kzalloc() with GFP_ATOMIC in atomic context

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

After the change, the post-commit hook output the following message:
"CHECK: Prefer kzalloc(sizeof(*pcmd_obj)...) over kzalloc(sizeof(struct cmd_obj)...)".

According to the above "CHECK", use the preferred style in the first kzalloc().

---

v1->v2: Fix an overlooked error due to an incorrect copy-paste
of the sizeof() operator.

 drivers/staging/r8188eu/core/rtw_mlme_ext.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/r8188eu/core/rtw_mlme_ext.c +2 -2
index 55c3d4a6faeb..315902682292 100644
@@ -6845,12 +6845,12 @@ void report_del_sta_event(struct adapter *padapter, unsigned char *MacAddr, unsi
 	struct mlme_ext_priv		*pmlmeext = &padapter->mlmeextpriv;
 	struct cmd_priv *pcmdpriv = &padapter->cmdpriv;
 
-	pcmd_obj = kzalloc(sizeof(struct cmd_obj), GFP_KERNEL);
+	pcmd_obj = kzalloc(sizeof(*pcmd_obj), GFP_ATOMIC);
 	if (!pcmd_obj)
 		return;
 
 	cmdsz = (sizeof(struct stadel_event) + sizeof(struct C2HEvent_Header));
-	pevtcmd = kzalloc(cmdsz, GFP_KERNEL);
+	pevtcmd = kzalloc(cmdsz, GFP_ATOMIC);
 	if (!pevtcmd) {
 		kfree(pcmd_obj);
 		return;
Signature · 1 line
2.33.1
Acked-by on v2 1/1
#1 lore raw
Quoted 13 lines · Fabio M. De Francesco

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

After the change, the post-commit hook output the following message:
"CHECK: Prefer kzalloc(sizeof(*pcmd_obj)...) over kzalloc(sizeof(struct cmd_obj)...)".

Quoted 8 lines

According to the above "CHECK", use the preferred style in the first kzalloc().

---

v1->v2: Fix an overlooked error due to an incorrect copy-paste
of the sizeof() operator.

I am happy that you caught the error before it destroyed every instance of r8188eu. Incidentally, I disagree with checkpatch in that I think that sizeof(struct foo) is more descriptive than sizeof(*bar). If I wanted to check the resulting value of the sizeof(), the second form requires an additional step. It probably does not matter much to the compiler, but when I have to do it manually, the extra effort is not negligible.

Even though I disagree with the philosophy,

Quoted 24 lines
  drivers/staging/r8188eu/core/rtw_mlme_ext.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/r8188eu/core/rtw_mlme_ext.c +2 -2
index 55c3d4a6faeb..315902682292 100644
@@ -6845,12 +6845,12 @@ void report_del_sta_event(struct adapter *padapter, unsigned char *MacAddr, unsi
  	struct mlme_ext_priv		*pmlmeext = &padapter->mlmeextpriv;
  	struct cmd_priv *pcmdpriv = &padapter->cmdpriv;
 
-	pcmd_obj = kzalloc(sizeof(struct cmd_obj), GFP_KERNEL);
+	pcmd_obj = kzalloc(sizeof(*pcmd_obj), GFP_ATOMIC);
  	if (!pcmd_obj)
  		return;
 
  	cmdsz = (sizeof(struct stadel_event) + sizeof(struct C2HEvent_Header));
-	pevtcmd = kzalloc(cmdsz, GFP_KERNEL);
+	pevtcmd = kzalloc(cmdsz, GFP_ATOMIC);
  	if (!pevtcmd) {
  		kfree(pcmd_obj);
  		return;
#2 lore raw
Quoted 6 lines · CET Larry Finger
Quoted 5 lines · Fabio M. De Francesco

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

Quoted 8 lines
Quoted 6 lines

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

[]

I am happy that you caught the error before it destroyed every instance of r8188eu.

I don't think so, since we have run this driver with no problems at all :)

SAC bugs can potentially cause serious system hangs at runtime, but they do not always cause problems in real execution as you have noticed here with this driver. We have used and tested it hundreds of times with no problems.

Quoted 2 lines

Incidentally, I disagree with checkpatch in that I think that sizeof(struct foo) is more descriptive than sizeof(*bar).

I agree with you in full, but I felt that I had to change it just because of the warning output by that tool. I don't like to have my patches discarded because they don't fix checkpatch warnings or introduce new ones.

Quoted 3 lines

If I wanted to check
the resulting value of the sizeof(), the second form requires an additional step. It probably does not matter much to the compiler, but when I have to

do it

Quoted 6 lines

manually, the extra effort is not negligible.

Even though I disagree with the philosophy,

Thanks for your "Acked-by" tag.

Fabio

#3 lore raw
Quoted 4 lines · Fabio M. De Francesco
Quoted 2 lines · CET Larry Finger

Incidentally, I disagree with checkpatch in that I think that sizeof(struct foo) is more descriptive than sizeof(*bar).

I agree with you in full

It's not checkpatch in particular, it's from coding-style

The preferred form for passing a size of a struct is the following:

.. code-block:: c

	p = kmalloc(sizeof(*p), ...);
#4 lore raw
In reply to Fabio M. De Francesco
Quoted 13 lines · Fabio M. De Francesco

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

After the change, the post-commit hook output the following message:
"CHECK: Prefer kzalloc(sizeof(*pcmd_obj)...) over kzalloc(sizeof(struct cmd_obj)...)".

Quoted 8 lines

According to the above "CHECK", use the preferred style in the first kzalloc().

---

v1->v2: Fix an overlooked error due to an incorrect copy-paste
of the sizeof() operator.

What commit does this fix?

thanks,

greg k-h

#5 lore raw
Quoted 20 lines · CET Greg Kroah-Hartman
Quoted 19 lines · Fabio M. De Francesco

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

After the change, the post-commit hook output the following message:
"CHECK: Prefer kzalloc(sizeof(*pcmd_obj)...) over kzalloc(sizeof(struct cmd_obj)...)".

According to the above "CHECK", use the preferred style in the first kzalloc().

---

Quoted 9 lines
Quoted 2 lines

v1->v2: Fix an overlooked error due to an incorrect copy-paste
of the sizeof() operator.

What commit does this fix?

thanks,

greg k-h

Sorry, Greg. Please let me know if I understand correctly what you are asking for...

In v1 I introduced a silly error while copy-pasting "sizeof()" and then I fixed it in v2.

I think that you mean that I should reword the list of changes from v1 because I'm not explaining properly why I submitted v2.

Is my understanding correct? If so, I have no problem in submitting v3.

Thank you in advance,

Fabio

P.S.: I had to resend this email and I want apologize for the noise. It seems that it contained HTML parts and for this reasons it was rejected by the relevant lists.

#6 lore raw
Quoted 33 lines · Fabio M. De Francesco
Quoted 20 lines · CET Greg Kroah-Hartman
Quoted 19 lines · Fabio M. De Francesco

Use the GFP_ATOMIC flag of kzalloc() with two memory allocation in report_del_sta_event(). This function is called while holding spinlocks, therefore it is not allowed to sleep. With the GFP_ATOMIC type flag, the allocation is high priority and must not sleep.

This issue is detected by Smatch which emits the following warning:
"drivers/staging/r8188eu/core/rtw_mlme_ext.c:6848 report_del_sta_event()
warn: sleeping in atomic context".

After the change, the post-commit hook output the following message:
"CHECK: Prefer kzalloc(sizeof(*pcmd_obj)...) over kzalloc(sizeof(struct cmd_obj)...)".

According to the above "CHECK", use the preferred style in the first kzalloc().

---

Quoted 9 lines
Quoted 2 lines

v1->v2: Fix an overlooked error due to an incorrect copy-paste
of the sizeof() operator.

What commit does this fix?

thanks,

greg k-h

Sorry, Greg. Please let me know if I understand correctly what you are asking for...

Quoted 7 lines

In v1 I introduced a silly error while copy-pasting "sizeof()" and then I fixed it in v2.

I think that you mean that I should reword the list of changes from v1 because I'm not explaining properly why I submitted v2.

Is my understanding correct? If so, I have no problem in submitting v3.

Sorry, no, I mean what commit in the kernel tree is this patch "fixing"?

You should have a "Fixes: " tag in the signed-off-by area of the changelog so that we know where the problem you are resolving here originated.

thanks,

greg k-h