On 10/05/2021 18:28, Ferruh Yigit wrote:
> On 5/10/2021 6:04 PM, Wang, Haiyue wrote:
>>> -----Original Message-----
>>> From: dev <dev-boun...@dpdk.org> On Behalf Of Ferruh Yigit
>>> Sent: Monday, May 10, 2021 23:03
>>> To: Yang, Qiming <qiming.y...@intel.com>; Zhang, Qi Z
>>> <qi.z.zh...@intel.com>; Stillwell Jr, Paul M
>>> <paul.m.stillwell...@intel.com>; Lu, Wenzhuo <wenzhuo...@intel.com>; Rong,
>>> Leyi <leyi.r...@intel.com>;
>>> Shukla, Shivanshu <shivanshu.shu...@intel.com>
>>> Cc: Yigit, Ferruh <ferruh.yi...@intel.com>; dev@dpdk.org; sta...@dpdk.org;
>>> Kevin Traynor
>>> <ktray...@redhat.com>; Ajit Khaparde <ajit.khapa...@broadcom.com>
>>> Subject: [dpdk-dev] [PATCH 3/4] net/ice/base: fix build with gcc11
>>>
>>> Reproduced with '--buildtype=debugoptimized' config,
>>> compiler version: gcc (GCC) 12.0.0 20210509 (experimental)
>>>
>>> There are multiple build errors, like:
>>> ../drivers/net/ice/base/ice_switch.c: In function ‘ice_add_marker_act’:
>>> ../drivers/net/ice/base/ice_switch.c:3727:15:
>>> warning: array subscript ‘struct ice_aqc_sw_rules_elem[0]’
>>> is partly outside array bounds of ‘unsigned char[52]’
>>> [-Warray-bounds]
>>> 3727 | lg_act->type = CPU_TO_LE16(ICE_AQC_SW_RULES_T_LG_ACT);
>>> | ^~
>>> In file included from ../drivers/net/ice/base/ice_type.h:52,
>>> from ../drivers/net/ice/base/ice_common.h:8,
>>> from ../drivers/net/ice/base/ice_switch.h:8,
>>> from ../drivers/net/ice/base/ice_switch.c:5:
>>> ../drivers/net/ice/base/ice_osdep.h:209:29:
>>> note: referencing an object of size 52 allocated by ‘rte_zmalloc’
>>> 209 | #define ice_malloc(h, s) rte_zmalloc(NULL, s, 0)
>>> | ^~~~~~~~~~~~~~~~~~~~~~~
>>> ../drivers/net/ice/base/ice_switch.c:3720:50:
>>> note: in expansion of macro ‘ice_malloc’
>>> lg_act = (struct ice_aqc_sw_rules_elem *)ice_malloc(hw, rules_size);
>>>
>>> These errors are mainly because allocated memory is cast to
>>> "struct ice_aqc_sw_rules_elem *" but allocated size is less than the size
>>> of "struct ice_aqc_sw_rules_elem".
>>>
>>> "struct ice_aqc_sw_rules_elem" has multiple other structs has unions,
>>> based on which one is used allocated memory being less than the size of
>>> "struct ice_aqc_sw_rules_elem" is logically correct but compiler is
>>> complaining about it.
>>>
>>> As a solution making sure allocated memory size is at least size of
>>> "struct ice_aqc_sw_rules_elem".
>>> The function to use the struct is 'ice_aq_sw_rules()', and it already has
>>> parameter for size of the rule, allocating more than needed shouldn't
>>> cause any problem.
>>>
Bugzilla ID: 678
>>> Fixes: c7dd15931183 ("net/ice/base: add virtual switch code")
>>> Fixes: 02acdce2f553 ("net/ice/base: add MAC filter with marker and counter")
>>> Fixes: f89aa3affa9e ("net/ice/base: support removing advanced rule")
>>> Cc: sta...@dpdk.org
>>>
If I apply on head of dpdk-next-net on F34 I still get some of these
warnings in ice. I'm using 'meson --werror -Dtests=false build'.
$ gcc --version
gcc (GCC) 11.1.1 20210428 (Red Hat 11.1.1-1)
>>> Signed-off-by: Ferruh Yigit <ferruh.yi...@intel.com>
>>> ---
>>> Cc: paul.m.stillwell...@intel.com
>>> Cc: qi.z.zh...@intel.com
>>> Cc: leyi.r...@intel.com
>>> Cc: Kevin Traynor <ktray...@redhat.com>
>>> Cc: Ajit Khaparde <ajit.khapa...@broadcom.com>
>>> ---
>>> drivers/net/ice/base/ice_switch.c | 30 +++++++++++++++++++++++-------
>>> 1 file changed, 23 insertions(+), 7 deletions(-)
>>
>> GCC bug ?
>>
>> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98266
>>
>> Bug 98266 - [11 Regression] bogus array subscript is partly outside array
>> bounds on virtual inheritance
This gcc defect is fixed in PR middle-end/98266 which was first in in
11.0.1-0.2 [1]. Currently Fedora 34 uses 11.1.1-1 and includes this PR.
[1] https://src.fedoraproject.org/rpms/gcc/blob/rawhide/f/gcc.spec#_3246
>>
> > I am not sure if this is a gcc defect.
>
> Here there is a memory allocated and assigned to "struct ice_aqc_sw_rules_elem
> *", but allocated memory size is less than the struct size. As far as I
> understand this is the reason of compiler warning.
>
> For this case it may not be problem logically since both who allocates memory
> and who uses the memory follows a contract, but there is a mismatch between
> pointer type and object. If some other function wants to access all fields of
> the struct, it will be out of bound access.
>
>
>
>
