public inbox for gcc-cvs@sourceware.org help / color / mirror / Atom feed
From: Eric Botcazou <ebotcazou@gcc.gnu.org> To: gcc-cvs@gcc.gnu.org Subject: [gcc r11-10463] Fix problematic interaction between bitfields, unions, SSO and SRA Date: Wed, 11 Jan 2023 19:47:11 +0000 (GMT) [thread overview] Message-ID: <20230111194711.782143858C83@sourceware.org> (raw) https://gcc.gnu.org/g:01e80c4c630a5a6286a521b047d0ef80631c892c commit r11-10463-g01e80c4c630a5a6286a521b047d0ef80631c892c Author: Eric Botcazou <ebotcazou@adacore.com> Date: Wed Jan 11 15:58:47 2023 +0100 Fix problematic interaction between bitfields, unions, SSO and SRA The handling of bitfields by the SRA pass is peculiar and this must be taken into account to support the scalar_storage_order attribute. gcc/ PR tree-optimization/108199 * tree-sra.c (sra_modify_expr): Deal with reverse storage order for bit-field references. gcc/testsuite/ * gcc.dg/sso-17.c: New test. Diff: --- gcc/testsuite/gcc.dg/sso-17.c | 52 +++++++++++++++++++++++++++++++++++++++++++ gcc/tree-sra.c | 18 ++++++++++++++- 2 files changed, 69 insertions(+), 1 deletion(-) diff --git a/gcc/testsuite/gcc.dg/sso-17.c b/gcc/testsuite/gcc.dg/sso-17.c new file mode 100644 index 00000000000..67e2d3793fd --- /dev/null +++ b/gcc/testsuite/gcc.dg/sso-17.c @@ -0,0 +1,52 @@ +/* { dg-do run } */ +/* { dg-options "-O2" } */ + +#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ +#define REV_ENDIANNESS __attribute__((scalar_storage_order("big-endian"))) +#else +#define REV_ENDIANNESS __attribute__((scalar_storage_order("little-endian"))) +#endif + +typedef unsigned long long u64; + +union DST { + u64 val; + + struct { + u64 x : 1; + u64 y : 1; + u64 r: 62; + } REV_ENDIANNESS; +} REV_ENDIANNESS; + + +struct SRC { + u64 a; +} REV_ENDIANNESS; + +[[gnu::noipa]] +void foo () {__builtin_abort();} + +[[gnu::noinline]] +int bar(struct SRC *src) +{ + union DST dst; + + dst.val = src->a; + + if (dst.y) { + foo(); + } + return 0; +} + +int main(void) +{ +#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ + struct SRC t = {-1ull & (~(0x01ull<<62))}; +#else + struct SRC t = {-1ull & (~(0x01ull<<1))}; +#endif + bar(&t); + return 0; +} diff --git a/gcc/tree-sra.c b/gcc/tree-sra.c index dea9cf3235c..5a1fc81d301 100644 --- a/gcc/tree-sra.c +++ b/gcc/tree-sra.c @@ -3828,7 +3828,23 @@ sra_modify_expr (tree *expr, gimple_stmt_iterator *gsi, bool write) } } else - *expr = repl; + { + /* If we are going to replace a scalar field in a structure with + reverse storage order by a stand-alone scalar, we are going to + effectively byte-swap the scalar and we also need to byte-swap + the portion of it represented by the bit-field. */ + if (bfr && REF_REVERSE_STORAGE_ORDER (bfr)) + { + REF_REVERSE_STORAGE_ORDER (bfr) = 0; + TREE_OPERAND (bfr, 2) + = size_binop (MINUS_EXPR, TYPE_SIZE (TREE_TYPE (repl)), + size_binop (PLUS_EXPR, TREE_OPERAND (bfr, 1), + TREE_OPERAND (bfr, 2))); + } + + *expr = repl; + } + sra_stats.exprs++; } else if (write && access->grp_to_be_debug_replaced)
reply other threads:[~2023-01-11 19:47 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=20230111194711.782143858C83@sourceware.org \ --to=ebotcazou@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).