On Wed, Jan 22, 2020 at 1:16 PM Andrew Pinski <pins...@gmail.com> wrote: > > On Wed, Jan 22, 2020 at 12:48 PM Jeff Law <l...@redhat.com> wrote: > > > > On Mon, 2020-01-20 at 09:42 +0100, Richard Biener wrote: > > > On Sat, Jan 18, 2020 at 1:47 AM <apin...@marvell.com> wrote: > > > > From: Andrew Pinski <apin...@marvell.com> > > > > > > > > On MIPS, .set noreorder/reorder needs to emitted around > > > > the nop. The template for the nop instruction uses %(/%) to > > > > do that. But default_print_patchable_function_entry uses > > > > fprintf rather than output_asm_insn to output the instruction. > > > > > > > > This fixes the problem by using output_asm_insn to emit the nop > > > > instruction. > > > > > > > > OK? Bootstrapped and tested on x86_64-linux-gnu and built a full > > > > mips toolchain also. > > > > > > OK. > > FWIW, I think this may have broke the arc-elf port. I'm getting > > failures for the patchable function entry tests. It looks like the > > port wants to peek a the current_output_insn in its handling of an > > output punctuation characters and current_output_insn is NULL. > > I suspect arc-elf was failing beforehand; just not crashing the compiler :). > Before this patch we would be printing out "nop%?" for arc-elf. The > tests are "compile" so they would have "passed" but only because the > tests was not trying to assemble them. If someone had tried to use > this option of arc-elf, they would have ran into a similar problem as > mips, printing out %? (in arc case).
Just a quick survey of the targets which had an issue before this patch: * arc - uses %? * bpf - uses %% * mips - uses %(/%) which emits .set noreorder/reorder * nios2 - %. * s390 - %% nios2 - checks current_output_insn for null-ness when it is seeing if it needs to print . or not. arc - does not check current_output_insn for null-ness Thanks, Andrew Pinski > > Thanks, > Andrew Pinski > > > > > > jeff > > > > >
