* [AArch64/GCC][0/N] Refactor prologue/epilogue code and optimize insns generated
@ 2014-07-22 14:49 Jiong Wang
0 siblings, 0 replies; only message in thread
From: Jiong Wang @ 2014-07-22 14:49 UTC (permalink / raw)
To: gcc-patches
currently, the implementation of AArch64 prologue/epilogue expand
hooks are a bit unclean.
the major issues are:
* when pushing callee-saved registers, register offset are
calculated by ad-hoc code instead of referencing offset
table initialized in aarch64_layout_frame.
* FP/LR push/restore are done by ad-hoc code, should be
unified with other registers.
* various duplicated code for pushing/restoring core and
vectore registers.
* sub-optimal instruction sequences generated for some
scenarios.
the following patch set trying to address above issues in
an incremental way, that the review could be easier.
no regression on aarch64-none-elf little/big endian bare-metal full test.
bootstrap OK on aarch64.
ok to install?
thanks.
-- Jiong
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2014-07-22 14:49 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-07-22 14:49 [AArch64/GCC][0/N] Refactor prologue/epilogue code and optimize insns generated Jiong Wang
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 read-only IMAP folder(s) and NNTP newsgroup(s).