On Wed, Apr 18, 2018 at 4:45 AM, Luc Van Oostenryck
<luc.vanoostenryck@xxxxxxxxx> wrote:
> On Wed, Apr 18, 2018 at 11:26:48AM +0200, Aurélien Aptel wrote:
>> Just forwarding a question I saw on linux-fsdevel:
>>
>> | From: Steve French <smfrench@xxxxxxxxx>
>> | Date: Wed, 4 Apr 2018 18:05:04 -0500
>> | Subject: Noisy Sparse warnings
>> | To: LKML <linux-kernel@xxxxxxxxxxxxxxx>,
>> | linux-fsdevel <linux-fsdevel@xxxxxxxxxxxxxxx>
>> | Sender: linux-fsdevel-owner@xxxxxxxxxxxxxxx
>> | Precedence: bulk
>> | List-ID: <linux-fsdevel.vger.kernel.org>
>> | X-Mailing-List: linux-fsdevel@xxxxxxxxxxxxxxx
>> |
>> | I have been getting the following sparse warnings repeatedly on my
>> | current Ubutu (17.10) when compiling the current mainline kernel:
>> |
>> | ./arch/x86/include/asm/nospec-branch.h:144:38: warning: Unknown escape '@'
>> | ./include/linux/string.h:239:1: error: attribute 'gnu_inline': unknown attribute
>> |
>> | I get the same sparse error whether I use the default sparse that
>> | Ubuntu ships (0.5.0) or the current sparse (0.5.2-rc1)
>> |
>> | Ideas?
>
> Thanks for the report.
>
> For the second warning, this will indeed happen with sparse-0.5.0 but
> have been fixed later. It certainly shouldn't happen with sparse-0.5.1
> (where sparse have been teached about 'gnu_inline') and even more so
> with 0.5.2-rc1 where such warnings have been disabled by default.
>
> For the first warning, it looks as a problem which have also been
> fixed before sparse-0.5.1.
>
> @Steve,
>
> Can you check that you're really using a recent sparse like 0.5.1 or
> 0.5.2-rc1?
Looks like 0.5.2-rc1 helped with two but now am getting a new one.
I have two systems - both are now verified (sparse --version) as running
version 0.5.2-rc1 for the user doing the build (that was part of the
problem, it was finding sparse 0.5.0 when running as a different user than
the one who had installed version 0.5.2-rc1) but I do get one error repeatedly
on one of the two Ubuntu systems (identical, current version, of Ubuntu)
CHECK fs/cifs/connect.c
fs/cifs/connect.c:825:25: warning: expression using sizeof(void)
See one of the dozen's of cases below:
cifs_dump_mem("Bad SMB: ", buf,
min_t(unsigned int, server->total_read, 48));
it looks like "min_t" is causing confusion for sparse (at least on the VM
running Ubuntu, the same version running on my laptop didn't generate
that when building with sparse).
--
Thanks,
Steve
--
To unsubscribe from this list: send the line "unsubscribe linux-sparse" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
