[lkml]   [2011]   [Mar]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH 04/15] ftrace: pack event structures.
On Fri, 2010-12-03 at 16:13 -0800, David Sharp wrote:
> Ftrace event structures have a 12-byte struct trace_entry at the beginning.
> If the structure is aligned, this means that if the first field is 64-bits,
> there will be 4 bytes of padding. Ironically, due to the 4-byte ringbuffer
> header, this will make 64-bit writes unaligned, if the ring buffer position
> is currently 64-bit aligned:
> 4(rb)+12(ftrace)+4(pad) = 20; 20%8 = 4

Actually the better answer is: It's time to nuke the lock_depth field.
That was added temporarily in helping to remove the BKL, and now that
it's pretty much gone, I say we need to nuke it. That will remove 4
bytes from the 12 byte ftrace header giving us a 8 byte header. Much
nicer :)

I'll do that today. Thanks!

-- Steve

> Adding __attribute__((packed)) to the event structures removes the extra
> space from the trace events, and actually improves alignment of trace
> events with a first field that is 64-bits.
> About 65 tracepoints have a 4-byte pad at offset 12:
> # find events -name format | xargs -n1 awk '
> $1=="name:" {name=$2}
> $1=="format:"{FS="\t"}
> $3=="offset:12;" && $4=="size:4;"{okay=1}
> $3=="offset:16;" && !okay {print name}' | wc -l
> 65
> With all 'syscalls' and 'timer' events enabled, this results in a 5%
> improvement in a simple 512MB read benchmark with warm caches.

 \ /
  Last update: 2011-03-09 14:05    [W:1.250 / U:0.736 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site