* fix: performance problem in crashReporter.start() on macOS
This change reduces the duration of crashReporter.start() on Intel macOS
from 622 milliseconds to 257 milliseconds!
Backports https://chromium-review.googlesource.com/c/crashpad/crashpad/+/3641386
posix: Replace DoubleForkAndExec() with ForkAndSpawn()
The DoubleForkAndExec() function was taking over 622 milliseconds to run
on macOS 11 (BigSur) on Intel i5-1038NG7. I did some debugging by adding
some custom traces and found that the fork() syscall is the bottleneck
here, i.e., the first fork() takes around 359 milliseconds and the
nested fork() takes around 263 milliseconds. Replacing the nested fork()
and exec() with posix_spawn() reduces the time consumption to 257
milliseconds!
See https://github.com/libuv/libuv/pull/3064 to know why fork() is so
slow on macOS and why posix_spawn() is a better replacement.
Another point to note is that even base::LaunchProcess() from Chromium
calls posix_spawnp() on macOS -
8f8d82dea0:base/process/launch_mac.cc;l=295-296
Change-Id: I25c6ee9629a1ae5d0c32b361b56a1ce0b4b0fd26
Reviewed-on: https://chromium-review.googlesource.com/c/crashpad/crashpad/+/3641386
Reviewed-by: Mark Mentovai <mark@chromium.org>
Commit-Queue: Mark Mentovai <mark@chromium.org>
Fixes: https://github.com/electron/electron/issues/34321
Signed-off-by: Darshan Sen <raisinten@gmail.com>
* Update .patches
Co-authored-by: Darshan Sen <raisinten@gmail.com>
Co-authored-by: Cheng Zhao <zcbenz@gmail.com>
* fix: crash when renderer process is reused
Could occur when a renderer crashes and the same-origin URL is loaded again
which leads to reusing the renderer process.
* test: renderer process crash recovery
* fix: handle case which leads to render frame DCHECK
* fix: lint
Co-authored-by: samuelmaddock <samuel.maddock@gmail.com>
* ci: cache python install to better deal with download errors.
* chore: use our CDN to download python2
* build: DRY up the python install steps
* fixup backport for 19-x-y
* trigger ci
* fixup the fixup
Co-authored-by: John Kleinschmidt <jkleinsc@electronjs.org>
Co-authored-by: Samuel Attard <sattard@salesforce.com>
Co-authored-by: Cheng Zhao <zcbenz@gmail.com>
chore: add a TRACE call for crash_reporter::Start()
Initializing the crashReporter takes around 620 milliseconds on Intel
macOS. I have sent a CL to crashpad to partially fix the performance
issue in
https://chromium-review.googlesource.com/c/crashpad/crashpad/+/3641386.
It would be beneficial to log the performance impact of this function in
the traces in case this slows down in the future.
Signed-off-by: Darshan Sen <raisinten@gmail.com>
Co-authored-by: Darshan Sen <raisinten@gmail.com>
reference: Note: It is known that having both Content-Security-Policy
and X-Content-Security-Policy or X-Webkit-CSP causes unexpected
behaviours on certain versions of browsers. Please avoid using deprecated
X-* headers. https://content-security-policy.com/
also:
1ad18486ed
Co-authored-by: Letu Ren <fantasquex@gmail.com>
fix: revert "feat: add first-instance-ack event to the `app.requestSingleInstanceLock()` flow (#34295)
* Revert "fix: requestSingleInstanceLock API ConnectNamedPipe sometimes hangs program (#33778)"
This reverts commit ffb8749243.
* fix: revert "feat: add first-instance-ack event to the `app.requestSingleInstanceLock()` flow"
* chore: update patch
