Re: [PATCH AArch64 2/2] Replace temporary inline assembler for vget_high
Ah, right you are. (Curiously I still can't find Richard's previous message in Thunderbird!) for first patch, gcc/testsuite/ChangeLog: * gcc.target/aarch64/vget_high_1.c: New test. * gcc.target/aarch64/vget_low_1.c: Likewise. for second patch, gcc/ChangeLog: * config/aarch64/arm_neon.h (__GET_HIGH): New macro. (vget_high_f32, vget_high_f64, vget_high_p8, vget_high_p16, vget_high_s8, vget_high_s16, vget_high_s32, vget_high_s64, vget_high_u8, vget_high_u16, vget_high_u32, vget_high_u64): Remove temporary __asm__ and reimplement. --Alan Marcus Shawcroft wrote: On 12 August 2014 11:12, Alan Lawrence alan.lawre...@arm.com wrote: This patch replaces the current inline assembler for the vget_high intrinsics in arm_neon.h with a sequence of other calls, in a similar fashion to vget_low. Unlike the assembler, these are all transparent to the front-end, so should enable better optimization through the mid-end. Tested check-gcc and check-g++ and aarch64-none-elf and aarch64_be-none-elf (including new tests in previous patch!). I think we are still waiting on ChangeLogs for this and the related patch? /Marcus
Re: [PATCH AArch64 2/2] Replace temporary inline assembler for vget_high
On 4 September 2014 14:21, Alan Lawrence alan.lawre...@arm.com wrote: Ah, right you are. (Curiously I still can't find Richard's previous message in Thunderbird!) for first patch, gcc/testsuite/ChangeLog: * gcc.target/aarch64/vget_high_1.c: New test. * gcc.target/aarch64/vget_low_1.c: Likewise. for second patch, gcc/ChangeLog: * config/aarch64/arm_neon.h (__GET_HIGH): New macro. (vget_high_f32, vget_high_f64, vget_high_p8, vget_high_p16, vget_high_s8, vget_high_s16, vget_high_s32, vget_high_s64, vget_high_u8, vget_high_u16, vget_high_u32, vget_high_u64): Remove temporary __asm__ and reimplement. OK both. /Marcus
Re: [PATCH AArch64 2/2] Replace temporary inline assembler for vget_high
On 12 August 2014 11:12, Alan Lawrence alan.lawre...@arm.com wrote: This patch replaces the current inline assembler for the vget_high intrinsics in arm_neon.h with a sequence of other calls, in a similar fashion to vget_low. Unlike the assembler, these are all transparent to the front-end, so should enable better optimization through the mid-end. Tested check-gcc and check-g++ and aarch64-none-elf and aarch64_be-none-elf (including new tests in previous patch!). I think we are still waiting on ChangeLogs for this and the related patch? /Marcus