Re: lib/raid6: SSSE3 optimized recovery functions v2
From: "H. Peter Anvin" <hpa@zytor.com>
Date: 2012-04-12 20:14:19
Also in:
lkml
On 04/12/2012 01:04 PM, Jim Kukunas wrote:
This keeps with the existing code conventions. The code block is: extern const u8 raid6_gfmul[256][256] __attribute__((aligned(256))); extern const u8 raid6_vgfmul[256][32] __attribute__((aligned(256))); extern const u8 raid6_gfexp[256] __attribute__((aligned(256))); extern const u8 raid6_gfinv[256] __attribute__((aligned(256))); extern const u8 raid6_gfexi[256] __attribute__((aligned(256)));quoted
WARNING: printk() should include KERN_ facility level #120: FILE: lib/raid6/algos.c:103: + printk("raid6: using %s recovery algorith\n", nest->name); WARNING: printk() should include KERN_ facility level #122: FILE: lib/raid6/algos.c:105: + printk("raid6: Yikes! No recovery algorithm found!\n"); WARNING: printk() should include KERN_ facility level #159: FILE: lib/raid6/algos.c:176: + printk("raid6: Yikes! No memory available.\n");Again, these are following the conventions of the existing code such as: printk("raid6: using algorithm %s (%ld MB/s)\n", In fact, the last printk, about no memory available, was simply moved to a different line in my patch.
A lot of the RAID-6 code predates the modern kernel conventions. It would be good to clean that up, but that is largely orthogonal to this patch. -hpa