Re: [PATCH] wireless: ath: work around false-positive stringop-overread warning

From: Kalle Valo <kvalo@kernel.org>
To: "Arnd Bergmann" <arnd@arndb.de>
Cc: "Arnd Bergmann" <arnd@kernel.org>,
	"David S . Miller" <davem@davemloft.net>,
	"Eric Dumazet" <edumazet@google.com>,
	"Jakub Kicinski" <kuba@kernel.org>,
	"Paolo Abeni" <pabeni@redhat.com>,
	"Johannes Berg" <johannes.berg@intel.com>,
	"Manikanta Pubbisetty" <quic_mpubbise@quicinc.com>,
	"Wen Gong" <quic_wgong@quicinc.com>,
	"Baochen Qiang" <quic_bqiang@quicinc.com>,
	"Sowmiya Sree Elavalagan" <quic_ssreeela@quicinc.com>,
	ath11k@lists.infradead.org, linux-wireless@vger.kernel.org,
	Netdev <netdev@vger.kernel.org>,
	linux-kernel@vger.kernel.org, ath12k@lists.infradead.org
Subject: Re: [PATCH] wireless: ath: work around false-positive stringop-overread warning
Date: Mon, 08 May 2023 17:57:39 +0300	[thread overview]
Message-ID: <87mt2eoopo.fsf@kernel.org> (raw)
In-Reply-To: <504c5a7d-0bfd-4b1e-a7f0-65d072657e0a@app.fastmail.com> (Arnd Bergmann's message of "Mon, 08 May 2023 10:52:11 +0200")

"Arnd Bergmann" <arnd@arndb.de> writes:

> On Mon, May 8, 2023, at 10:44, Kalle Valo wrote:
>
>> Arnd Bergmann <arnd@kernel.org> writes:
>>
>>> From: Arnd Bergmann <arnd@arndb.de>
>>>
>>> In a rare arm64 randconfig build, I got multiple warnings for ath11k
>>> and ath12k:
>>>
>>> In function 'ath11k_peer_assoc_h_ht',
>>>     inlined from 'ath11k_peer_assoc_prepare' at drivers/net/wireless/ath/ath11k/mac.c:2665:2:
>>> drivers/net/wireless/ath/ath11k/mac.c:1709:13: error: 'ath11k_peer_assoc_h_ht_masked' reading 10 bytes from a region of size 0 [-Werror=stringop-overread]
>>>  1709 |         if (ath11k_peer_assoc_h_ht_masked(ht_mcs_mask))
>>>       |             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>
>>> This happens whenever gcc-13 fails to inline one of the functions
>>> that take a fixed-length array argument but gets passed a pointer.
>>>
>>> Change these functions to all take a regular pointer argument
>>> instead.
>>>
>>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>>
>> s/wireless:/wifi:/ but I can fix that.
>
> Ok, thanks!
>
>> In a awat it's a shame to lose the explicit length but I guess there's
>> no other way to fix this?
>
> There might be, but I couldn't figure out a way that works.

Ok.

>> Also I hope you find the time to add GCC 13 to crosstool :) Related to
>> this
>
> I uploaded gcc-13.1.0 binaries last week, but still need to
> update the html page, so it's not yet linked. You can navigate
> the directories from the gcc-12 builds.

Thanks! I was able to find the build[1] but having an issue:

$ ./x86_64-linux-gcc -v
./x86_64-linux-gcc: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.35' not found (required by ./x86_64-linux-gcc)
./x86_64-linux-gcc: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.32' not found (required by ./x86_64-linux-gcc)
./x86_64-linux-gcc: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by ./x86_64-linux-gcc)
./x86_64-linux-gcc: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.36' not found (required by ./x86_64-linux-gcc)
./x86_64-linux-gcc: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.34' not found (required by ./x86_64-linux-gcc)

With older GCC versions from your page I don't have this problem. I'm
using Debian 10 still so so is my libc too old?

ii  libc6:amd64          2.28-10+deb10u2:amd6 GNU C Library: Shared libraries

[1] https://mirrors.edge.kernel.org/pub/tools/crosstool/files/bin/x86_64/13.1.0/x86_64-gcc-13.1.0-nolibc-x86_64-linux.tar.gz

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches