netdev - Re: [Intel-wired-lan] Is this an out-of-bounds issue?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <b7dabc2d-19a4-42f4-ba5c-31e1803d821a@intel.com>
Date: Mon, 16 Sep 2024 11:19:32 +0200
From: Przemek Kitszel <przemyslaw.kitszel@...el.com>
To: Qianqiang Liu <qianqiang.liu@....com>, Ahmed Zaki <ahmed.zaki@...el.com>
CC: <intel-wired-lan@...ts.osuosl.org>, <anthony.l.nguyen@...el.com>,
	<linux-kernel@...r.kernel.org>, <netdev@...r.kernel.org>
Subject: Re: [Intel-wired-lan] Is this an out-of-bounds issue?

On 9/16/24 10:10, Przemek Kitszel wrote:
> On 9/13/24 15:13, Qianqiang Liu wrote:
>> The code in drivers/net/ethernet/intel/ice/ice_parser_rt.c:
>>
>> 114 static void ice_bst_key_init(struct ice_parser_rt *rt,
>> 115                              struct ice_imem_item *imem)
>> 116 {
>> 117         u8 tsr = (u8)rt->gpr[ICE_GPR_TSR_IDX];
>> 118         u16 ho = rt->gpr[ICE_GPR_HO_IDX];
>> 119         u8 *key = rt->bst_key;
>> 120         int idd, i;
>> 121
>> 122         idd = ICE_BST_TCAM_KEY_SIZE - 1;
>> 123         if (imem->b_kb.tsr_ctrl)
>> 124                 key[idd] = tsr;
>> 125         else
>> 126                 key[idd] = imem->b_kb.prio;
>>
>> The "ICE_BST_TCAM_KEY_SIZE" macro is 20, so "idd" is 20 - 1 = 19.
>> "key" equals "rt->bst_key" which is an array, and the size of the
>> array is ICE_BST_KEY_SIZE which is 10.
>> Is it possible that 'key[idd]' might access invalid memory?
>> Should the "idd" be "ICE_BST_KEY_SIZE"?
>>
>> -    idd = ICE_BST_TCAM_KEY_SIZE - 1;
>> +    idd = ICE_BST_KEY_SIZE - 1;
>>
> 
> We already have a fix for that from Ahmed, but it is not yet public.
> @Ahmed, please follow up.
> 

ugh, sorry, it's already public:
https://lore.kernel.org/all/20240823230847.172295-1-ahmed.zaki@intel.com/

awaits our VAL