1454745 - Skeletal bootstrap process

Status: RESOLVED FIXED

(Firefox :: General, enhancement, P2)

Description

[(No longer employed by Mozilla) Aaron Klotz]

See also https://bugzilla.mozilla.org/show_bug.cgi?id=1435780#c1

This bug should just involve adding the command-line flag and basic infra for doing a bootstrap. We'll add the actual policies etc in follow-up bugs.

Comment 1

[(No longer employed by Mozilla) Aaron Klotz]

Attachment: Skeletal bootstrap process

This is the just the essentials that allow us to test bootstrapping by specifying --launcher on the command line. We'll add policies and blocklist setup in follow-up bugs.

Comment 2

[(No longer employed by Mozilla) Aaron Klotz]

Comment on attachment 8972147 [details] [diff] [review]
Skeletal bootstrap process

Moving this to Matt as per PlatInt meeting today.

Comment 3

[Molly Howell (she/her) (mostly inactive)]

Comment on attachment 8972147 [details] [diff] [review]
Skeletal bootstrap process

Review of attachment 8972147 [details] [diff] [review]:
-----------------------------------------------------------------

Yep, looks good. Just a couple small points.

I thought about maybe adding some tests for this patch, but I feel like we're good without any, since we'll have plenty of testing of how the bootstrap process is working out when it's made the default behavior, and anyway I'm not sure how to verify "passing a particular command-line flag has no observable effect", which is what the test for this patch would boil down to.

::: browser/app/LauncherProcessWin.cpp
@@ +57,5 @@
> +  if (!result) {
> +    return;
> +  }
> +
> +  ::MessageBoxW(nullptr, rawMsgBuf, L"Firefox", MB_OK | MB_ICONERROR);

I don't love hard-coding "Firefox" here, but I don't have a good way to avoid it, so I'll have to live with it.

@@ +69,5 @@
> +bool
> +RunAsLauncherProcess(int& argc, wchar_t** argv)
> +{
> +  return CheckArg(argc, argv, L"launcher",
> +                  reinterpret_cast<const wchar_t**>(nullptr),

I tried to build this patch in Clang, and it wouldn't compile this cast; apparently the language spec doesn't allow reinterpret_casts from nullptr. It seems to only want to take a static_cast.

Comment 4

[(No longer employed by Mozilla) Aaron Klotz]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=f099499b1d193d800bc4f19522d119f48e485365

Comment 5

[(No longer employed by Mozilla) Aaron Klotz]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=35af29dc46ae0cc92d41a228557da69b3fe27364

Comment 6

[(No longer employed by Mozilla) Aaron Klotz]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=848a95b3ffa8ff84304d934c9174358b521f7349

Comment 7

[(No longer employed by Mozilla) Aaron Klotz]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=386455145d86b9079e9a46544f26557637786250

Comment 8

[(No longer employed by Mozilla) Aaron Klotz]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=b6958dc379b0bbb19ee0b56bcbe32d927c9f253e

Comment 9

[(No longer employed by Mozilla) Aaron Klotz]

https://hg.mozilla.org/integration/mozilla-inbound/rev/eb5c30217df94a8f2f48b6cce953ff717cb3be89
Bug 1454745: Skeletal bootstrap process; r=mhowell

Comment 10

[Arthur Iakab [arthur_iakab]]

https://hg.mozilla.org/mozilla-central/rev/eb5c30217df9