| Message ID | 5755A061.1000909@foss.arm.com |
|---|---|
| State | New |
| Headers | show |
On 06/06/16 17:10, Kyrill Tkachov wrote: > Hi all, > > This small patch adds handling of the CSEL-type instructions to the Cortex-A57 scheduling model. > It is treated the same as simple ALU instructions. > > With this patch I didn't see any overall differences in SPEC2006. > > Bootstrapped and tested on arm-none-linux-gnueabihf and aarch64-linux-gnu. > > Ok for trunk? > > The patch is very simple and the csel value isn't used in any arm instructions so I think > just an aarch64 approval for this should be enough. > > Thanks, > Kyrill > > 2016-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com> > > * config/arm/cortex-a57.md (cortex_a57_alu): > Handle csel type. This is ok by me - in these sorts of situations I think we should let the aarch64 approvals be sufficient. regards Ramana
On Tue, Jun 07, 2016 at 09:22:05AM +0100, Ramana Radhakrishnan wrote: > > > On 06/06/16 17:10, Kyrill Tkachov wrote: > > Hi all, > > > > This small patch adds handling of the CSEL-type instructions to the Cortex-A57 scheduling model. > > It is treated the same as simple ALU instructions. > > > > With this patch I didn't see any overall differences in SPEC2006. > > > > Bootstrapped and tested on arm-none-linux-gnueabihf and aarch64-linux-gnu. > > > > Ok for trunk? > > > > The patch is very simple and the csel value isn't used in any arm instructions so I think > > just an aarch64 approval for this should be enough. > > > > Thanks, > > Kyrill > > > > This is ok by me - in these sorts of situations I think we should let the > aarch64 approvals be sufficient. I agree. I don't think the folder in which the file resides is the most important thing, it is the intent and risk of the patch that matters. This patch is OK for AArch64. Thanks, James > > 2016-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com> > > > > * config/arm/cortex-a57.md (cortex_a57_alu): > > Handle csel type. >
diff --git a/gcc/config/arm/cortex-a57.md b/gcc/config/arm/cortex-a57.md index 37912db464315a0d70835b81991e8e07a4d9db89..c8cf80f4ba7ed99b46c920c2d0ad3299050ec473 100644 --- a/gcc/config/arm/cortex-a57.md +++ b/gcc/config/arm/cortex-a57.md @@ -297,7 +297,7 @@ (define_insn_reservation "cortex_a57_alu" 2 (eq_attr "type" "alu_imm,alus_imm,logic_imm,logics_imm,\ alu_sreg,alus_sreg,logic_reg,logics_reg,\ adc_imm,adcs_imm,adc_reg,adcs_reg,\ - adr,bfm,clz,rbit,rev,alu_dsp_reg,\ + adr,bfm,clz,csel,rbit,rev,alu_dsp_reg,\ rotate_imm,shift_imm,shift_reg,\ mov_imm,mov_reg,\ mvn_imm,mvn_reg,\
Hi all, This small patch adds handling of the CSEL-type instructions to the Cortex-A57 scheduling model. It is treated the same as simple ALU instructions. With this patch I didn't see any overall differences in SPEC2006. Bootstrapped and tested on arm-none-linux-gnueabihf and aarch64-linux-gnu. Ok for trunk? The patch is very simple and the csel value isn't used in any arm instructions so I think just an aarch64 approval for this should be enough. Thanks, Kyrill 2016-06-06 Kyrylo Tkachov <kyrylo.tkachov@arm.com> * config/arm/cortex-a57.md (cortex_a57_alu): Handle csel type.