Fix MIPS tests

[eduardosm]

Tests failed to build due to use of legacy const generics syntax, out-of-bounds immediates and incorrect name of const generics in assert_instr

[rustbot]

r? @sayantn

rustbot has assigned @sayantn.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: @Amanieu, @folkertdev, @sayantn
• @Amanieu, @folkertdev, @sayantn expanded to Amanieu, folkertdev, sayantn
• Random selection from Amanieu, folkertdev, sayantn

[folkertdev]

View changes since the review

what's happening here?

[folkertdev]

ah 3 == 67 % 64, and this function uses static_assert_uimm_bits!(IMM5, 5);

[sayantn]

For some reason I can't run the mips tests, ci/run-docker.sh mips64-unknown-linux-gnuabi64 runs into a few errors. E.g. the static_assert_simm_bits assertions all fail (even with this patch, and the failures don't make any sense, e.g it says 5 is not between -17 and 16)

[folkertdev]

I think some of the assert logic might be buggy (across targets)?

https://godbolt.org/z/qP74MWvK1

in any case the bit_width version expresses what we mean more accurately.

[folkertdev]

Ah, it's signed and that is weird because you need a bit for the sign. Definitely a poor error message in any case.

[eduardosm]

I had checked assert_instr tests with cargo check --release, which does not check the static asserts. Then I ran tests with cargo test (debug mode), which does not compile assert_instr at all, so I missed those failures.

Many assert_instr are failing anyways, and I will not be fixing them for now.

[eduardosm]

Windows job failing, unrelated I believe

  = note: C:/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/15.2.0/../../../../x86_64-w64-mingw32/bin/ld.exe: D:\a\stdarch\stdarch\target\x86_64-pc-windows-gnu\debug\deps\stdarch_gen_arm-c2d4a4c482dd3e81.9ul22rteqt6x41b1oy83npkbi.1831grd.rcgu.o:9ul22rteqt6x41b1oy83npkbi:(.debug_gdb_scripts+0x0): multiple definition of `.weak.__rustc_debug_gdb_scripts_section__.default'; D:\a\stdarch\stdarch\target\x86_64-pc-windows-gnu\debug\deps\stdarch_gen_arm-c2d4a4c482dd3e81.1nzalzaaigcyndp95lkud74oo.1831grd.rcgu.o:1nzalzaaigcyndp95lkud74oo:(.debug_gdb_scripts+0x0): first defined here

[folkertdev]

yes it's unrelated rust-lang/rust#155277

[folkertdev]

Many assert_instr are failing anyways, and I will not be fixing them for now.

Fair enough. Did anything in particular cause you to look at this?

(the changes look fine to me, so I think we should just merge this when CI works again with the next nightly)

[eduardosm]

Did anything in particular cause you to look at this?

I was going to continue reducing the use of unsafe fn in simd_test functions (#1981, #1987, #1989, #2006), but I rather have tests compiling before making more changes.

[folkertdev]

We removed the CI jobs for these targets, but probably should just still have them but make them not blocking for PRs? Then at least we know when they are broken and can occasionally fix them.