Fingerprinting Protections v2: Workers - Canvas - follow up to 11368

[LaurenWags]

Description

Found while testing #11368

Per discussion with @pes10k, assigning to @pilgrim-brave for investigation.

Per instructions on dev-pages.brave.software/farbling.html and dev-pages.bravesoftware.com/farbling.html, for Canvas the "Worker" column value should match the other columns for that row when FP = standard or strict (since Mode has "both" for this row) and the value should be different between the two pages.

This ONLY occurs on dev-pages.bravesoftware.com/farbling.html, whereas dev-pages.brave.software/farbling.html works as expected.

Steps to Reproduce

1. Clean profile, 1.16.x
2. Navigate to dev-pages.brave.software/farbling.html and dev-pages.bravesoftware.com/farbling.html
3. Confirm FP = standard (or strict, either one will reproduce the issue)
4. Generate Fingerprints

Actual result:

FP = standard

FP = strict

Expected result:

Worker column value should match other values in the same row (but should be different between the two pages)

Reproduces how often:

easily

Desktop Brave version:

Brave	1.16.58 Chromium: 86.0.4240.75 (Official Build) dev (x86_64)
Revision	c69c33933bfc72a159aceb4aeca939eb0087416c-refs/branch-heads/4240@{#1149}
OS	macOS Version 10.14.6 (Build 18G3020)

Android Device details:

• Install type (ARM, x86):
• Device type (Phone, Tablet, Phablet):
• Android version:

Version/Channel Information:

• Can you reproduce this issue with the current release? n/a
• Can you reproduce this issue with the beta channel? yes 1.16.x
• Can you reproduce this issue with the nightly channel? yes 1.17.x

Other Additional Information:

• Does the issue resolve itself when disabling Brave Shields?
• Does the issue resolve itself when disabling Brave Rewards?
• Is the issue reproducible on the latest version of Chrome?

Miscellaneous Information:

cc @rebron

[LaurenWags]

Reproduces in 1.17.x as well:

FP = standard

FP = strict

[pes10k]

I believe this is fixed now. @LaurenWags can you check again when you have a chance?

[LaurenWags]

@pes10k same question 😄 - what version do you think this should be checked against - Nightly (currently 1.21.x) or one of the other channels?

[pes10k]

Ah, sorry @LaurenWags ! It should be fixed in nightly / 1.21.x

[pes10k]

ah, apologies @LaurenWags , i think this one is not ready for another test (i jumped the gun).. But the other, WebGL one should be good to go. Sorry for the confusion!

[LaurenWags]

no worries, will give the other one a try and put some notes on it. Thanks!

[pes10k]

Closing as this is not a bug (its a result of the canvas API using a different path, and so a different intervention point, than main frame farbling)