On Mon, 20 Mar 2023 at 20:19, Pavel Machek <pavel@xxxxxx> wrote: > > Hi! > > > [ Upstream commit 8849818679478933dd1d9718741f4daa3f4e8b86 ] > > > > The kernel disables all SSE and similar FP/SIMD instructions on > > x86-based architectures (partly because we shouldn't be using floats in > > the kernel, and partly to avoid the need for stack alignment, see: > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53383 ) > > > > UML does not do the same thing, which isn't in itself a problem, but > > does add to the list of differences between UML and "normal" x86 builds. > > > > In addition, there was a crash bug with LLVM < 15 / rustc < 1.65 when > > building with SSE, so disabling it fixes rust builds with earlier > > compiler versions, see: > > https://github.com/Rust-for-Linux/linux/pull/881 > > We don't have rust in 4.14, so cited problem can't hit us. This should > not go to -stable. > > (Plus, KBUILD_RUSTFLAGS is not going to exist in -stable). > I agree, this is not a good fit for -stable. While I'd argue the KBUILD_CFLAGS part is still technically valid: - As noted, the KBUILD_RUSTFLAGS bit is useless without Rust support. - It triggers a bug in older gcc versions (< 11), which is bad anyway, and probably worse for -stable: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99652 There's a patch to work around the latter: https://lore.kernel.org/linux-um/20230318041555.4192172-1-davidgow@xxxxxxxxxx/ But I'd agree that excluding this change from -stable altogether is the better option. Cheers, -- David > Best regards, > Pavel > > > +# Disable SSE and other FP/SIMD instructions to match normal x86 > > +# > > +KBUILD_CFLAGS += -mno-sse -mno-mmx -mno-sse2 -mno-3dnow -mno-avx > > +KBUILD_RUSTFLAGS += -Ctarget-feature=-sse,-sse2,-sse3,-ssse3,-sse4.1,-sse4.2,-avx,-avx2 > > + > > ifeq ($(CONFIG_X86_32),y) > > START := 0x8048000 > > > > -- > People of Russia, stop Putin before his war on Ukraine escalates.
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature
