fuchsia-binary-size increase in CL 5607565: IDB: Add an inspect control in indexeddb-internals

[Abhishek Shanthkumar]

Hi fuchsia-dev, my CL from the subject line failed the fuchsia-binary-size tryjob with a message saying the uncompressed size of web_engine grew by 4096 bytes (and the compressed size by 16384 bytes) - failed - fuchsia-binary-size 767455 | LUCI (chromium.org).

Can you please take an initial look at the CL and the tryjob result and confirm that this is not a false positive and needs to be investigated?

I browsed through the Chrome Speed - Trybot: fuchsia-binary-size doc and will try to run the check locally to figure out which exact part of the CL caused the size increase.

Do I need a Linux/Mac machine running on ARM64 to build Chrome-Fuchsia, or can I do it on an x64-based Windows machine too? I'm guessing I can't, but thought I'll check...

Many thanks for your help!

Regards,

Abhishek

 

[Abhishek Shanthkumar]

Never mind, the tryjob passed on retry. Do let me know if you see anything of concern in the first run though!

Thanks,

Abhishek

---

From: Abhishek Shanthkumar
Sent: Thursday, June 13, 2024 7:04:59 PM
To: fuchs...@chromium.org <fuchs...@chromium.org>
Cc: est...@chromium.org <est...@chromium.org>; Steve Becker <ste...@microsoft.com>
Subject: fuchsia-binary-size increase in CL 5607565: IDB: Add an inspect control in indexeddb-internals

 

[Zijie He]

Thank you Abhishek for reaching out. This is a known issue, though not very common. The compression logic may not guarantee the size increment of the compressed data would always be smaller than the uncompressed data.

The trybot itself creates an error if the increment of the compressed data is over 16K. Since it is based on the diff, rebasing the baseline, modifying the change, etc. may impact the results. Though usually it should be deterministic, if not, something may be wrong and worth an extra investigation.

For your change, 4K uncompressed size increment is OK, unfortunately it caused a 16K compressed size increment and triggered an error.

If a later run is failing with the same symptom, i.e. 4K uncompressed and 16K compressed, you may follow https://chromium.googlesource.com/chromium/src/+/main/docs/speed/binary_size/fuchsia_binary_size_trybot.md#if-all-else-fails to ignore the error with “Fuchsia-Binary-Size: $ANY_TEXT_HERE” in the change description.

--
You received this message because you are subscribed to the Google Groups "fuchsia-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fuchsia-dev...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/fuchsia-dev/MN2PR00MB08611FECB6A3A9BD5D3E48A890C12%40MN2PR00MB0861.namprd00.prod.outlook.com.

[Abhishek Shanthkumar]

Got it, thank you for the detailed response!

Thanks,

Abhishek

 

---

From: Zijie He <zij...@chromium.org>
Sent: Thursday, June 13, 2024 10:46 PM
To: Abhishek Shanthkumar <Abhishek.S...@microsoft.com>
Cc: fuchs...@chromium.org <fuchs...@chromium.org>; est...@chromium.org <est...@chromium.org>; Steve Becker <ste...@microsoft.com>
Subject: [EXTERNAL] Re: fuchsia-binary-size increase in CL 5607565: IDB: Add an inspect control in indexeddb-internals

 

You don't often get email from zij...@chromium.org. Learn why this is important