| 12 Sep 2026 |
 K900 | Might want to ask the Nobara folks | 11:07:50 |
| K900 | Aaaaand they're gone | 11:08:02 |
 nyanbinary 🏳️⚧️ | Should JSGC_DISABLE_POISONING=1 be enabled in the wrapper maybe for perf (I mean this only matters for nightly really tbf) | 19:59:02 |
| nyanbinary 🏳️⚧️ | Also did not know Firefox had a --extended-corpus option for PGO | 19:59:22 |
| nyanbinary 🏳️⚧️ | --set-default should be fine since users can override it anyways :p | 22:38:09 |
| nyanbinary 🏳️⚧️ | o nvm Mozilla just never updated their documentation saying this is off by default now for optimizied Nightly builds lol | 23:04:12 |
| 13 Sep 2026 |
| nyanbinary 🏳️⚧️ | hexa By the way is there a reason buildMozillaMach doesn't pass --enable-rust-simd like Mozilla does for their official builds of Firefox? | 02:16:41 |
 hexa | me sleep | 02:17:06 |
| hexa | and check back next week | 02:17:11 |
| nyanbinary 🏳️⚧️ | alright | 02:18:03 |
| nyanbinary 🏳️⚧️ | So you don't go out searching this is where Mozilla defines it:
https://github.com/mozilla-firefox/firefox/blob/main/build/mozconfig.rust
Then mozconfig.rust gets called by mozconfig.common here:
https://github.com/mozilla-firefox/firefox/blob/main/build/mozconfig.common | 02:18:41 |
| nyanbinary 🏳️⚧️ | https://github.com/NixOS/nixpkgs/blob/master/pkgs/build-support/build-mozilla-mach/default.nix Nixpkgs buildMozillaMach doesn't pass --enable-rust-simd here... | 02:19:13 |
| 14 Sep 2026 |
| nyanbinary 🏳️⚧️ | more work for you i suppose hexa
https://groups.google.com/a/mozilla.org/g/dev-platform/c/qlaQ1YSlOP8?pli=1 | 00:53:39 |
| nyanbinary 🏳️⚧️ | 2 weeks release cycles instead of 4 weeks for firefox :p | 00:53:54 |
| hexa | still the same week | 00:54:01 |
| hexa | tjat' | 00:54:10 |
| hexa | Redacted or Malformed Event | 00:54:13 |
| hexa | 156.0 tomorrow-ish | 00:54:18 |
| hexa | 155.0 two weeks ago | 00:54:22 |
| nyanbinary 🏳️⚧️ | icic | 00:55:46 |
| hexa | # SIMD acceleration for encoding_rs
# ==============================================================
option(
"--enable-rust-simd", env="MOZ_RUST_SIMD", help="Enable explicit SIMD in Rust code"
)
@depends("--enable-rust-simd", target)
def rust_simd(value, target):
# As of 2019-09-17, the simd-accel feature of encoding_rs has not
# been properly set up outside aarch64, armv7, x86 and x86_64.
if target.cpu in ("aarch64", "arm", "x86", "x86_64") and value:
return True
set_config("MOZ_RUST_SIMD", rust_simd)
set_define("MOZ_RUST_SIMD", rust_simd)
| 01:06:08 |
| hexa | 2019-2020 code | 01:06:53 |
| hexa | mildly exciting | 01:07:46 |
| nyanbinary 🏳️⚧️ | Meaning what :p, as far as I can tell in the code Mozilla still uses it so it probably still does something | 01:08:22 |
| hexa | not much except that the knob has been there a while now | 01:08:49 |
| nyanbinary 🏳️⚧️ | Older :p
2017
https://bugzilla.mozilla.org/show_bug.cgi?id=1261841 | 01:19:01 |
| nyanbinary 🏳️⚧️ | https://bugzilla.mozilla.org/show_bug.cgi?id=1261841#c176 also this person asked if downstream should enable it which they responded with yes | 01:19:59 |
| nyanbinary 🏳️⚧️ | Also mentioned implictly adding it to --enable-release which they never got around to doing so I guess lol | 01:20:33 |
| nyanbinary 🏳️⚧️ | * Also mentioned adding it to --enable-release which they never got around to doing so I guess lol | 01:29:59 |
| nyanbinary 🏳️⚧️ | Arch Linux uses --enable-rust-simd hmm
https://www.mail-archive.com/arch-commits%40lists.archlinux.org/msg1035514.html
https://www.mail-archive.com/arch-commits%40lists.archlinux.org/msg697468.html
So we just forgot to enable that and it just was overlooked until now | 01:41:55 |
