Revision 8f71465c19ffefbfd0da3c1f5dc172b4bce05e93 authored by Jeff Layton on 01 May 2012, 21:41:49 UTC, committed by Steve French on 02 May 2012, 03:27:54 UTC
While testing, I've found that even when we are able to negotiate a much larger rsize with the server, on-the-wire reads often end up being capped at 128k because of ra_pages being capped at that level. Lifting this restriction gave almost a twofold increase in sequential read performance on my craptactular KVM test rig with a 1M rsize. I think this is safe since the actual ra_pages that the VM requests is run through max_sane_readahead() prior to submitting the I/O. Under memory pressure we should end up with large readahead requests being suppressed anyway. Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
1 parent 156d179
File | Mode | Size |
---|---|---|
3com | ||
acenic | ||
adaptec | ||
advansys | ||
av7110 | ||
bnx2 | ||
bnx2x | ||
cis | ||
cpia2 | ||
cxgb3 | ||
dabusb | ||
dsp56k | ||
e100 | ||
edgeport | ||
emi26 | ||
emi62 | ||
ess | ||
kaweth | ||
keyspan | ||
keyspan_pda | ||
korg | ||
matrox | ||
myricom | ||
ositech | ||
qlogic | ||
r128 | ||
radeon | ||
sb16 | ||
sun | ||
tehuti | ||
tigon | ||
ttusb-budget | ||
vicam | ||
yam | ||
yamaha | ||
.gitignore | -rw-r--r-- | 39 bytes |
Makefile | -rw-r--r-- | 11.7 KB |
README.AddingFirmware | -rw-r--r-- | 1.5 KB |
WHENCE | -rw-r--r-- | 27.7 KB |
atmsar11.HEX | -rw-r--r-- | 18.7 KB |
ihex2fw.c | -rw-r--r-- | 6.6 KB |
intelliport2.bin.ihex | -rw-r--r-- | 92.2 KB |
mts_cdma.fw.ihex | -rw-r--r-- | 37.2 KB |
mts_edge.fw.ihex | -rw-r--r-- | 37.8 KB |
mts_gsm.fw.ihex | -rw-r--r-- | 37.2 KB |
ti_3410.fw.ihex | -rw-r--r-- | 37.0 KB |
ti_5052.fw.ihex | -rw-r--r-- | 37.0 KB |
tr_smctr.bin.ihex | -rw-r--r-- | 20.5 KB |
whiteheat.HEX | -rw-r--r-- | 43.9 KB |
whiteheat_loader.HEX | -rw-r--r-- | 11.8 KB |
whiteheat_loader_debug.HEX | -rw-r--r-- | 17.2 KB |
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...