[libvirt] [PATCH] maint: turn on gcc logical-op checking

Matthias Bolte matthias.bolte at googlemail.com
Wed Jul 28 13:37:45 UTC 2010


2010/7/26 Daniel Veillard <veillard at redhat.com>:
> On Sat, Jul 24, 2010 at 10:36:50PM +0200, Matthias Bolte wrote:
>> 2010/7/23 Eric Blake <eblake at redhat.com>:
> [...]
>> From: Eric Blake <eblake at redhat.com>
>> Date: Fri, 23 Jul 2010 14:28:31 -0600
>> Subject: [PATCH] maint: turn on gcc logical-op checking
>>
>> This would have detected the bug in commit 38ad33931 (Aug 09), which
>> we missed until commit f828ca35 (Jul 10); over 11 months later.
>>
>> However, on Fedora 13, it also triggers LOTS of warnings from
>> the libcurl-devel header for two files:
>>
>> esx/esx_vi.c: In function 'esxVI_CURL_Perform':
>> esx/esx_vi.c:232: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> esx/esx_vi.c:232: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> esx/esx_vi.c:232: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> ...
>> xenapi/xenapi_driver.c: In function 'call_func':
>> xenapi/xenapi_driver.c:1872: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> xenapi/xenapi_driver.c:1872: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> xenapi/xenapi_driver.c:1872: error: logical '&&' with non-zero constant will always evaluate as true [-Wlogical-op]
>> ...
>
>  ACK to Matthias' patch, is there is a predefined way to avoid those
>  defines, let's use it !
>
> Daniel
>

Thanks, pushed.

Matthias




More information about the libvir-list mailing list