Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 26 Dec 2022 13:03:59 -0800
From:   Kees Cook <kees@kernel.org>
To:     Linus Torvalds <torvalds@linux-foundation.org>,
        Guenter Roeck <linux@roeck-us.net>,
        Jaegeuk Kim <jaegeuk@kernel.org>, Chao Yu <chao@kernel.org>
CC:     Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Vlastimil Babka <vbabka@suse.cz>,
        Peter Zijlstra <peterz@infradead.org>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Kees Cook <keescook@chromium.org>
Subject: Re: Linux 6.2-rc1
User-Agent: K-9 Mail for Android
In-Reply-To: <CAHk-=whD1zMyt4c7g6-+tWvVweyb-6oHMT_+ZVHqe1EXwtFpCQ@mail.gmail.com>
References: <CAHk-=wgf929uGOVpiWALPyC7pv_9KbwB2EAvQ3C4woshZZ5zqQ@mail.gmail.com> <20221226195206.GA2626419@roeck-us.net> <CAHk-=whD1zMyt4c7g6-+tWvVweyb-6oHMT_+ZVHqe1EXwtFpCQ@mail.gmail.com>
Message-ID: <DA632860-284E-4923-8863-9D2745DD289E@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On December 26, 2022 12:56:29 PM PST, Linus Torvalds <torvalds@linux-founda=
tion=2Eorg> wrote:
>On Mon, Dec 26, 2022 at 11:52 AM Guenter Roeck <linux@roeck-us=2Enet> wro=
te:
>>
>> fs/f2fs/inline=2Ec: In function 'f2fs_move_inline_dirents':
>> include/linux/fortify-string=2Eh:59:33: error: '__builtin_memset' point=
er overflow between offset [28, 898293814] and size [-898293787, -1] [-Werr=
or=3Darray-bounds]
>> fs/f2fs/inline=2Ec:430:9: note: in expansion of macro 'memset'
>>   430 |         memset(dst=2Ebitmap + src=2Enr_bitmap, 0, dst=2Enr_bitm=
ap - src=2Enr_bitmap);
>>       |         ^~~~~~
>
>Well, that's unfortunate=2E

I'll look into this=2E

> [=2E=2E=2E]
>> Not exactly a regression, but worth mentioning:
>>
>> CONFIG_MEMCPY_KUNIT_TEST now sometimes takes several minutes to
>> execute in qemu=2E On top of that, it may result in hung task timeouts
>> if the hung task timeout is set to low values (45 seconds and below)=2E
>> Example, seen with s390:
>>
>> =2E=2E=2E
>> [   18=2E494320]     ok 2 memcpy_test
>> [   52=2E969037]     ok 3 memcpy_large_test
>> =2E=2E=2E
>> [   52=2E974505]     ok 4 memmove_test
>> [   87=2E325400]     ok 5 memmove_large_test
>> [  143=2E562760] INFO: task swapper/0:1 blocked for more than 46 second=
s=2E
>> =2E=2E=2E
>> [  143=2E564441] Call Trace:
>> [  143=2E564689]  [<0000000000f1ec80>] __schedule+0x370/0x720
>> [  143=2E565175]  [<0000000000f1f098>] schedule+0x68/0x110
>> [  143=2E565374]  [<0000000000f278d4>] schedule_timeout+0xc4/0x160
>> [  143=2E565603]  [<0000000000f1fde2>] __wait_for_common+0xda/0x250
>> [  143=2E565816]  [<0000000000903c90>] kunit_try_catch_run+0x98/0x178
>> [  143=2E566029]  [<0000000000f05c9c>] kunit_run_case_catch_errors+0x7c=
/0xb8
>> [  143=2E566956]  [<00000000009023c0>] kunit_run_tests+0x220/0x638
>> =2E=2E=2E
>>
>> That is too much for my test bed=2E I dropped this test as result=2E Th=
is means
>> that extending the tests has, at least in the context of my testing, th=
e
>> opposite effect=2E
>
>Kees? This indeed seems counter-productive=2E=2E

Hrm, that is not supposed to take THAT long=2E=2E=2E But yes a cross-arxh =
qemu run would be slow=2E :(

The changes there were to help find any future memcpy problem (like seen w=
hile porting the i386 memcpy asm=2E=2E=2E) I'll try to get this reduced=2E =
Dropping the test isn't so great=2E :)


--=20
Kees Cook