Tags: andrunko/cobalt
Tags
Cleanup comments left over from youtube#4830 (youtube#4883) This cleans up two lines of commented out code leftover from verifying the localhost test.
Roll Chrome Mac PGO Profile Roll Chrome Mac PGO profile from chrome-mac-6834-1737199987-cf126707c82f5811411a05c3b67761208d1a264f-10453ec4d397daef9dddd60eee920dee4770762f.profdata to chrome-mac-6834-1737222910-aaa30023f1e92973c5cfeb5dce94419351198550-83929666c97d2053ca9d0948c7af06eb00303d70.profdata If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/pgo-mac-chromium-stable Please CC chrome-brapp-engprod@google.com,pgo-profile-sheriffs@google.com on the revert to ensure that a human is aware of the problem. To file a bug in Chromium stable branch: https://bugs.chromium.org/p/chromium/issues/entry To report a problem with the AutoRoller itself, please file a bug: https://issues.skia.org/issues/new?component=1389291&template=1850622 Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md Cq-Include-Trybots: luci.chrome-m132.try:mac-chrome Tbr: pgo-profile-sheriffs@google.com Change-Id: I4fc1df10ebc8091264999d69fc62d08490b10959 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6183615 Commit-Queue: Chrome Release Autoroll <chromium-release-autoroll@skia-public.iam.gserviceaccount.com> Bot-Commit: Chrome Release Autoroll <chromium-release-autoroll@skia-public.iam.gserviceaccount.com> Cr-Commit-Position: refs/branch-heads/6834@{youtube#3956} Cr-Branched-From: 47a3549fac11ee8cb7be6606001ede605b302b9f-refs/heads/main@{#1381561}
[M126-LTS][VideoCaptureManager] Replace raw pointers with scoped_refptr VCM used VideoCaptureController raw pointers in a number of places, including as a field in VCM::CaptureDeviceStartRequest. This CL replaces the field and some other usages with a scoped_refptr to prevent dangling pointers. (cherry picked from commit 3524ce528548d1d743a6aa6e339ecb5a186c22bc) Bug: 382135228 Change-Id: I1bd5f95bdf57631227034beb8bb076f258606378 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6088073 Commit-Queue: Guido Urdaneta <guidou@chromium.org> Reviewed-by: Dale Curtis <dalecurtis@chromium.org> Cr-Original-Commit-Position: refs/heads/main@{#1396301} Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6099792 Reviewed-by: Guido Urdaneta <guidou@chromium.org> Commit-Queue: Gyuyoung Kim (xWF) <qkim@google.com> Cr-Commit-Position: refs/branch-heads/6478@{youtube#2009} Cr-Branched-From: e6143acc03189c5e52959545b110d6d17ecd5286-refs/heads/main@{#1300313}
Explicitly set v8 thread priority (youtube#4693) b/378556308 b/386841458
Fix send for connected UDP socket (youtube#4402) In this PR [youtube#4270](youtube#4270), functionality was added to connect a UDP socket. When we call the `SbSocketSendTo` function, which invokes `sendto` for a UDP socket, the issue arises that if this function is called for a socket that was previously connected, it throws an `errno 56 (already connected)` error on certain platforms, such as `macOS` or `BSD`. To address this, we need to call the send function instead of `sendto`. This behavior is platform-specific, as it works fine on Windows or Linux. To resolve this issue, I propose a minimal patch that, depending on whether the UDP socket is already connected, will call either `sendto` or send accordingly. In my opinion, another approach could involve splitting the `SbSocketSendTo` function into `SbSocketSendTo` (specifically for non-connected UDP sockets) and `SbSocketSend` (for TCP and connected UDP sockets). However, this would require significant code changes. I suggest keeping `SbSocketSendTo` and handling the connected socket state as I’ve done in this PR. In the future, when we fully transition to the POSIX API, we can directly call the appropriate function (`send` or `sendto`) within the `UDPSocketStarboard` class, instead of using wrappers like `SbSocketSendTo`. This transition would also create a clearer interface, as currently, when we call `SbSocketSendTo` for an already connected socket, we pass an address that is ultimately unused. I look forward to hearing your feedback.
Add Omaha config IDs to configurator. (youtube#4401) b/277111513
Add Omaha config IDs to configurator. (youtube#4401) b/277111513
Read multiple packets per call for QUIC connections (youtube#4321) This adds an extension, implemented for linux and Android that gives access to `recvmmsg` to read multiple packets in a single system call, reducing CPU usage in the kernel and the system call overhead. This add another 15% CPU reduction for networking on top of the one for 25.lts.12. On the FHD cast dongle on wired network, download benchmark shows more than 40% higher bandwidths and more than 15% reduction in CPU usage. b/376351154 b/205134049 b/376350096
Automated Commit: LKGM 15662.118.0 for chromeos. Uploaded by https://ci.chromium.org/b/8738505171997425025 Change-Id: I97cb38c7db8029bc94e733819844b638921f1940 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5817034 Bot-Commit: ChromeOS bot <3su6n15k.default@developer.gserviceaccount.com> Commit-Queue: Yuta Hijikata <ythjkt@chromium.org> Cr-Commit-Position: refs/branch-heads/6099@{youtube#2064} Cr-Branched-From: e6ee4500f7d6549a9ac1354f8d056da49ef406be-refs/heads/main@{#1217362}
PreviousNext