This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [AArch64][PATCH 5/5] Use atomic load-operate instructions for update-fetch patterns.
- From: Andrew Pinski <pinskia at gmail dot com>
- To: Matthew Wahab <matthew dot wahab at foss dot arm dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 17 Sep 2015 11:51:34 -0700
- Subject: Re: [AArch64][PATCH 5/5] Use atomic load-operate instructions for update-fetch patterns.
- Authentication-results: sourceware.org; auth=none
- References: <55FAEC63 dot 7040403 at foss dot arm dot com> <55FAF059 dot 2030205 at foss dot arm dot com>
On Thu, Sep 17, 2015 at 9:54 AM, Matthew Wahab
<matthew.wahab@foss.arm.com> wrote:
> Hello,
>
> ARMv8.1 adds atomic swap and atomic load-operate instructions with
> optional memory ordering specifiers. This patch uses the ARMv8.1
> load-operate instructions to implement the atomic_<op>_fetch patterns.
>
> The approach is to use the atomic load-operate instruction to atomically
> load the data and update memory and then to use the loaded data to
> calculate the value that the instruction would have stored. The
> calculation attempts to mirror the operation of the atomic instruction.
> For example, atomic_and_fetch<mode> is implemented with an atomic
> load-bic so the result is also calculated using a BIC instruction.
>
> The general form of the code generated for an atomic_<op>_fetch, with
> destination D, source S, memory address A and memory order MO, depends
> on whether or not the operation is directly supported by the
> instruction. If <op> is one of PLUS, IOR or XOR, the code generated is:
>
> ld<opc><mo><sz> S, D, [A]
> <inst> D, D, S
> where
> <opc> is one {add, set, eor}
> <inst> is one of {add, orr, xor}
> <mo> is one of {'', 'a', 'l', 'al'} depending on memory order MO.
> <sz> is one of {'', 'b', 'h'} depending on the data size.
>
> If <op> is SUB, the code generated is:
>
> neg S, S
> ldadd<mo><sz> S, D, [A]
> add D, D, S
>
> If <op> is AND, the code generated is:
>
> not S, S
> ldclr<mo><sz> S, D, [A]
> bic D, S, S
>
> Any operation not in {PLUS, IOR, XOR, SUB, AND} is passed to the
> existing aarch64_split_atomic_op function, to implement the operation
> using sequences built with the ARMv8 load-exclusive/store-exclusive
> instructions
>
> Tested the series for aarch64-none-linux-gnu with native bootstrap and
> make check. Also tested for aarch64-none-elf with cross-compiled
> check-gcc on an ARMv8.1 emulator with +lse enabled by default.
Are you going to add some builtins for MIN/MAX support too?
Thanks,
Andrew Pinski
>
> Ok for trunk?
> Matthew
>
> 2015-09-17 Matthew Wahab <matthew.wahab@arm.com>
>
> * config/aarch64/aarch64-protos.h (aarch64_gen_atomic_ldop):
> Adjust declaration.
> * config/aarch64/aarch64.c (aarch64_emit_bic): New.
> (aarch64_gen_atomic_load_op): Adjust comment. Add parameter
> out_result. Update to support update-fetch operations.
> * config/aarch64/atomics.md (aarch64_atomic_exchange<mode>_lse):
> Adjust for change to aarch64_gen_atomic_ldop.
> (aarch64_atomic_<atomic_optab><mode>_lse): Likewise.
> (aarch64_atomic_fetch_<atomic_optab><mode>_lse): Likewise.
> (atomic_<atomic_optab>_fetch<mode>): Change to an expander.
> (aarch64_atomic_<atomic_optab>_fetch<mode>): New.
> (aarch64_atomic_<atomic_optab>_fetch<mode>_lse): New.
>
> gcc/testsuite
> 2015-09-17 Matthew Wahab <matthew.wahab@arm.com>
>
> * gcc.target/aarch64/atomic-inst-ldadd.c: Add tests for
> update-fetch operations.
> * gcc.target/aarch64/atomic-inst-ldlogic.c: Likewise.
>