Created attachment 52034 [details] reduced testcase Output: $ x86_64-pc-linux-gnu-gcc -Oz testcase.c $ ./a.out Aborted $ x86_64-pc-linux-gnu-gcc -v Using built-in specs. COLLECT_GCC=/repo/gcc-trunk/binary-latest-amd64/bin/x86_64-pc-linux-gnu-gcc COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r12-6072-20211220084313-g8d1e342b4af-checking-yes-rtl-df-extra-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/12.0.0/lto-wrapper Target: x86_64-pc-linux-gnu Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++ --enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra --with-cloog --with-ppl --with-isl --build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu --target=x86_64-pc-linux-gnu --with-ld=/usr/bin/x86_64-pc-linux-gnu-ld --with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch --prefix=/repo/gcc-trunk//binary-trunk-r12-6072-20211220084313-g8d1e342b4af-checking-yes-rtl-df-extra-amd64 Thread model: posix Supported LTO compression algorithms: zlib zstd gcc version 12.0.0 20211220 (experimental) (GCC)
Sorry. Mine. Bootstrapping and regression testing a fix now.
Also: int a; void foo (void) { a = -42; } which corrupts some random unrelated memory. For the TARGET_64BIT movsi case, it needs to be done only if the destination is a register rather than MEM. Mostly unrelated, I wonder if we shouldn't have a separate TYPE_PUSHPOP, decide in the (set (attr "type") and then just emit it and get e.g. the "length" attribute right for it.
Patch proposed https://gcc.gnu.org/pipermail/gcc-patches/2021-December/587258.html
(In reply to Roger Sayle from comment #3) > Patch proposed > https://gcc.gnu.org/pipermail/gcc-patches/2021-December/587258.html Thank you for the patch. Does it work correctly with the red zone? I am asking since the "mov" instruction normally doesn't affect stack, and I don't see any indication that the "push" alternative does. But I definitely do not have enough knowledge about the GCC internals.
The master branch has been updated by Roger Sayle <sayle@gcc.gnu.org>: https://gcc.gnu.org/g:ef26c151c14a87177d46fd3d725e7f82e040e89f commit r12-6106-gef26c151c14a87177d46fd3d725e7f82e040e89f Author: Roger Sayle <roger@nextmovesoftware.com> Date: Thu Dec 23 12:33:07 2021 +0000 x86: PR target/103773: Fix wrong-code with -Oz from pop to memory. This is a fix to PR target/103773 where -Oz shouldn't use push/pop on x86 to shrink writing small integer constants to memory. Instead clang uses "andl $0, mem" for writing zero, and "orl $-1, mem" when writing -1 to memory when using -Oz. This patch implements this via peephole2 where we can confirm that its ok to clobber the flags. 2021-12-23 Roger Sayle <roger@nextmovesoftware.com> Uroš Bizjak <ubizjak@gmail.com> gcc/ChangeLog PR target/103773 * config/i386/i386.md (*mov<mode>_and): New define_insn for writing a zero to memory using AND. (*mov<mode>_or): Extend to allow memory destination and HImode. (*movdi_internal): Remove -Oz push/pop optimization from here. (*movsi_internal): Likewise. (peephole2): Perform -Oz push/pop optimization here, only for register destinations, values other than zero, and in functions that don't used the red zone. (peephole2): With -Oz, convert writes of 0 or -1 to memory into their clobber forms, i.e. *mov<mode>_and and *mov<mode>_or resp. gcc/testsuite/ChangeLog PR target/103773 * gcc.target/i386/pr103773-2.c: New test case. * gcc.target/i386/pr103773.c: New test case.
This shoud now be fixed on mainline.