Hi,
See commit ef9209b642f ("staging: rtl8723bs: Fix indenting errors and an
off-by-one mistake in core/rtw_mlme_ext.c") for detail.
I don't know how can you make sure that line 3254 can be true in the for
loop. If the condition never satisfies, then there is an off-by-one
access in line 3267.
If you can prove it, then the patch is unnecessary.
Young
On 2018/11/27 16:15, Dan Carpenter wrote:
> The original code is OK.
>
> On Tue, Nov 27, 2018 at 07:29:07AM +0000, Yang Xiao wrote:
>> From: Young_X <YangX92@xxxxxxxxxxx>
>>
>> The error at line 3267 was the result of an off-by-one error in
>> a for loop in line 3253.
>> If condition in line 3254 never satisfies, then the value of
>> pstat->aid is NUM_STA+1. This will lead to out-of-bound access
>> in line 3267.
> It's best to avoid line numbers in the commit message unless you paste
> the actual code, because sometimes people will be using different line
> numbers from you.
>
>> Signed-off-by: Young_X <YangX92@xxxxxxxxxxx>
>> ---
>> drivers/staging/rtl8188eu/core/rtw_mlme_ext.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/staging/rtl8188eu/core/rtw_mlme_ext.c b/drivers/staging/rtl8188eu/core/rtw_mlme_ext.c
>> index 6790b840..0854adc 100644
>> --- a/drivers/staging/rtl8188eu/core/rtw_mlme_ext.c
>> +++ b/drivers/staging/rtl8188eu/core/rtw_mlme_ext.c
>> @@ -3250,7 +3250,7 @@ static unsigned int OnAssocReq(struct adapter *padapter,
>> if (pstat->aid > 0) {
>> DBG_88E(" old AID %d\n", pstat->aid);
>> } else {
>> - for (pstat->aid = 1; pstat->aid <= NUM_STA; pstat->aid++)
>> + for (pstat->aid = 1; pstat->aid < NUM_STA; pstat->aid++)
> drivers/staging/rtl8188eu/core/rtw_mlme_ext.c
> 3249 /* get a unique AID */
> 3250 if (pstat->aid > 0) {
> 3251 DBG_88E(" old AID %d\n", pstat->aid);
> 3252 } else {
> 3253 for (pstat->aid = 1; pstat->aid <= NUM_STA; pstat->aid++)
> 3254 if (pstapriv->sta_aid[pstat->aid - 1] == NULL)
> 3255 break;
> 3256
> 3257 /* if (pstat->aid > NUM_STA) { */
> ^^^^^^^^^^^^^^^^^^^^^^^^
> This comment is key. pstapriv->max_num_sta is always less than or equal
> to NUM_STA.
>
> 3258 if (pstat->aid > pstapriv->max_num_sta) {
> 3259 pstat->aid = 0;
> 3260
> 3261 DBG_88E(" no room for more AIDs\n");
> 3262
> 3263 status = WLAN_STATUS_AP_UNABLE_TO_HANDLE_NEW_STA;
> 3264
> 3265 goto OnAssocReqFail;
> 3266 } else {
> 3267 pstapriv->sta_aid[pstat->aid - 1] = pstat;
> ^^^^^^^^^^^^^^
> So this is fine.
>
> 3268 DBG_88E("allocate new AID=(%d)\n", pstat->aid);
> 3269 }
> 3270 }
> 3271
>
> regards,
> dan carpenter
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
