public inbox for gcc-cvs@sourceware.org help / color / mirror / Atom feed
From: Roger Sayle <sayle@gcc.gnu.org> To: gcc-cvs@gcc.gnu.org Subject: [gcc r14-1902] i386: Two minor tweaks to ix86_expand_move. Date: Sat, 17 Jun 2023 21:32:02 +0000 (GMT) [thread overview] Message-ID: <20230617213202.A4B7F3858D28@sourceware.org> (raw) https://gcc.gnu.org/g:96c3539f2a38134cb76d8ab2e924e0dc70b2ccbd commit r14-1902-g96c3539f2a38134cb76d8ab2e924e0dc70b2ccbd Author: Roger Sayle <roger@nextmovesoftware.com> Date: Sat Jun 17 22:28:40 2023 +0100 i386: Two minor tweaks to ix86_expand_move. This patch splits out two (independent) minor changes to i386-expand.cc's ix86_expand_move from a larger patch, given that it's better to review and commit these independent pieces separately from a more complex patch. The first change is to test for CONST_WIDE_INT_P before calling ix86_convert_const_wide_int_to_broadcast. Whilst stepping through this function in gdb, I was surprised that the code was continually jumping into this function with operands that obviously weren't appropriate. The second change is to generalize the optimization for efficiently moving a TImode value to V1TImode (via V2DImode), to cover all 128-bit vector modes. Hence for the test case: typedef unsigned long uv2di __attribute__ ((__vector_size__ (16))); uv2di foo2(__int128 x) { return (uv2di)x; } we'd previously move via memory with: foo2: movq %rdi, -24(%rsp) movq %rsi, -16(%rsp) movdqa -24(%rsp), %xmm0 ret with this patch we now generate with -O2 (the same as V1TImode): foo2: movq %rdi, %xmm0 movq %rsi, %xmm1 punpcklqdq %xmm1, %xmm0 ret and with -O2 -msse4 the even better: foo2: movq %rdi, %xmm0 pinsrq $1, %rsi, %xmm0 ret The new test case is unimaginatively called sse2-v1ti-mov-2.c given the original test case just for V1TI mode was called sse2-v1ti-mov-1.c. 2023-06-17 Roger Sayle <roger@nextmovesoftware.com> gcc/ChangeLog * config/i386/i386-expand.cc (ix86_expand_move): Check that OP1 is CONST_WIDE_INT_P before calling ix86_convert_wide_int_to_broadcast. Generalize special case for converting TImode to V1TImode to handle all 128-bit vector conversions. gcc/testsuite/ChangeLog * gcc.target/i386/sse2-v1ti-mov-2.c: New test case. Diff: --- gcc/config/i386/i386-expand.cc | 10 ++++++---- gcc/testsuite/gcc.target/i386/sse2-v1ti-mov-2.c | 16 ++++++++++++++++ 2 files changed, 22 insertions(+), 4 deletions(-) diff --git a/gcc/config/i386/i386-expand.cc b/gcc/config/i386/i386-expand.cc index def060ab562..6a28b6767a7 100644 --- a/gcc/config/i386/i386-expand.cc +++ b/gcc/config/i386/i386-expand.cc @@ -521,7 +521,8 @@ ix86_expand_move (machine_mode mode, rtx operands[]) return; } } - else if (GET_MODE_SIZE (mode) >= 16) + else if (CONST_WIDE_INT_P (op1) + && GET_MODE_SIZE (mode) >= 16) { rtx tmp = ix86_convert_const_wide_int_to_broadcast (GET_MODE (op0), op1); @@ -696,8 +697,9 @@ ix86_expand_vector_move (machine_mode mode, rtx operands[]) return; } - /* Special case TImode to V1TImode conversions, via V2DI. */ - if (mode == V1TImode + /* Special case TImode to 128-bit vector conversions via V2DI. */ + if (VECTOR_MODE_P (mode) + && GET_MODE_SIZE (mode) == 16 && SUBREG_P (op1) && GET_MODE (SUBREG_REG (op1)) == TImode && TARGET_64BIT && TARGET_SSE @@ -709,7 +711,7 @@ ix86_expand_vector_move (machine_mode mode, rtx operands[]) emit_move_insn (lo, gen_lowpart (DImode, SUBREG_REG (op1))); emit_move_insn (hi, gen_highpart (DImode, SUBREG_REG (op1))); emit_insn (gen_vec_concatv2di (tmp, lo, hi)); - emit_move_insn (op0, gen_lowpart (V1TImode, tmp)); + emit_move_insn (op0, gen_lowpart (mode, tmp)); return; } diff --git a/gcc/testsuite/gcc.target/i386/sse2-v1ti-mov-2.c b/gcc/testsuite/gcc.target/i386/sse2-v1ti-mov-2.c new file mode 100644 index 00000000000..7e890857678 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/sse2-v1ti-mov-2.c @@ -0,0 +1,16 @@ +/* { dg-do compile { target int128 } } */ +/* { dg-options "-O2 -msse2" } */ + +typedef unsigned __int128 uv1ti __attribute__ ((__vector_size__ (16))); +typedef unsigned long uv2di __attribute__ ((__vector_size__ (16))); +typedef unsigned int uv4si __attribute__ ((__vector_size__ (16))); +typedef unsigned short uv8hi __attribute__ ((__vector_size__ (16))); +typedef unsigned char uv16qi __attribute__ ((__vector_size__ (16))); + +uv1ti foo1(__int128 x) { return (uv1ti)x; } +uv2di foo2(__int128 x) { return (uv2di)x; } +uv4si foo4(__int128 x) { return (uv4si)x; } +uv8hi foo8(__int128 x) { return (uv8hi)x; } +uv16qi foo16(__int128 x) { return (uv16qi)x; } + +/* { dg-final { scan-assembler-not "%\[er\]sp" } } */
reply other threads:[~2023-06-17 21:32 UTC|newest] Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20230617213202.A4B7F3858D28@sourceware.org \ --to=sayle@gcc.gnu.org \ --cc=gcc-cvs@gcc.gnu.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
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).