QA Report about 2024-03-zksync-findings

[alex-ppg]

Hey @Bauchibred, appreciate the in-depth analysis of the QA report and your contribution to the PJQA process! I have evaluated all findings that infer a medium upgrade as follows:

• QA-02: While this is indeed a mistake, I believe a QA (L) severity is appropriate as the off-chain software of zkSync Era relies on events rather than return variables (as a return variable is accessible to the caller whilst events are accessible to all listeners). I will raise this concern with the zkSync Era team for a revisit but will retain a QA (L) ruling for now.
• QA-06: This is invalid as the cumulative balance values of the contract are at least equal to the totalSupply due to the token being a closed-circuit system (i.e. balances cannot increase from 0 without totalSupply being increased as well)
• QA-09: The Additional Note item is indeed correct in the sense that regardless of whether this finding is valid, the exhibit cannot be considered as a valid HM due to being out-of-scope. One of the duplicates from that contest is actually mine, and I understand the implications fully, however, I believe they have been adequately covered in the previous contest and this item is out-of-scope.
• QA-13: Likewise, this is a valid QA finding from the perspective of additional insight into a past finding but cannot be considered in scope as an HM vulnerability.
• QA-28: I believe a QA (L) severity is fair, and it is impossible to mimic the extcodecopy instruction's operation due to zkSync Era's internal structure.

I believe that the present ruling is fair, but will make sure to notify the sponsor for a re-evaluation of QA-02 in case it merits an upgrade. It definitely is a mistake in the code, but I do not believe its ramifications to be impactful as nodes utilize events as highlighted in other exhibits such as #112.

[DecentralDisco]

Regarding validity of items in this QA report, per conversation with the judge @alex-ppg:

The only clear mistake is QA-06, and the rest are passable NC / L / I recommendations. As such, I confirm that all QA items except for QA-06 are valid.

As such, QA-06 will be excluded from the final audit report.

[razzorsec]

The first “QA” was considered medium by us!

[c4-judge]

alex-ppg marked the issue as grade-a

[c4-judge]

alex-ppg marked the issue as selected for report

[alex-ppg]

This submission was graded as the best due to exceeding the 50% accuracy threshold whilst containing valid and thoroughly elaborated findings in an easily digestible format. To note, the first QA finding will be upgraded accordingly once judging concludes and this exhibit is split into a second one as a duplicate of #97.

[Bauchibred]

Hi @alex-ppg, thanks for judging the contest, I'd like to ask if it's possible you have any comments in regards to the upgradability of some of the listed borderline low/medium issues, a couple were attached here in the QA report as an attempt on not spamming the judging repo with reports that could end up being finalised as QA, I'd appreciate a quick glance on this borderline issues to see if any could be upgraded.

To ease the re-review, I believe grepping the markdown file with the word medium would pinpoint most of these issues, however I'd appreciate re-review as not all have been linked with the medium word, thanks once again for your time.