Bug 80361

Summary:	metag/kernel/cachepart.c:102: bad way to test for failure ?
Product:	Platform Specific/Hardware	Reporter:	David Binderman (dcb314)
Component:	Other	Assignee:	platform_other
Status:	NEW ---
Severity:	normal	CC:	andrey_utkin
Priority:	P1
Hardware:	Other
OS:	Linux
Kernel Version:	3.16-rc5	Subsystem:
Regression:	No	Bisected commit-id:

Description David Binderman 2014-07-16 17:05:32 UTC

[linux-3.16-rc5/arch/metag/kernel/cachepart.c:102]: (style) Checking if unsigned variable 'thread_cache_size' is less than zero.

Source code is

        if (thread_cache_size < 0)
            pr_emerg("Can't read %s cache size\n",
                 cache_type ? "DCACHE" : "ICACHE");

but

    unsigned int thread_cache_size;

Function get_thread_cache_size returns an error code
as (unsigned int) -1. 

Maybe it would be better to test for that ?

Comment 1 Andrey Utkin 2014-07-17 13:31:18 UTC

Thanks, patch submitted: https://patchwork.kernel.org/patch/4575011/