* [Qemu-devel] Dynamic condition code calculation for Sparc
@ 2009-05-10 9:07 Blue Swirl
0 siblings, 0 replies; only message in thread
From: Blue Swirl @ 2009-05-10 9:07 UTC (permalink / raw)
To: qemu-devel
[-- Attachment #1: Type: text/plain, Size: 615 bytes --]
Hi,
I just committed a patch series, which implement dynamic condition
code calculation for Sparc, much like x86. The speedup for Sparc64 is
clearly visible, less so for Sparc32.
The attached follow-on optimization patches implement leaner carry
calculation and calculating the (annulled) branch brcond directly from
comparison sources, but currently they cause breakage.
Also the normal branches should be converted to calculate the brcond
directly, but this fails if the branch delay slot contains an
instruction that changes the flags. I had some code for this but had
to scratch it. Any suggestions welcome.
[-- Attachment #2: 0001-Branch-optimization-BROKEN.patch --]
[-- Type: application/mbox, Size: 4403 bytes --]
[-- Attachment #3: 0001-Convert-C-flag-input-BROKEN.patch --]
[-- Type: application/mbox, Size: 5476 bytes --]
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2009-05-10 9:07 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-05-10 9:07 [Qemu-devel] Dynamic condition code calculation for Sparc Blue Swirl
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).