netdev - Re: [TEST] net/bind_wildcard flakiness

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20250516174435.70447-1-kuniyu@amazon.com>
Date: Fri, 16 May 2025 10:44:31 -0700
From: Kuniyuki Iwashima <kuniyu@...zon.com>
To: <kuba@...nel.org>
CC: <kuniyu@...zon.com>, <netdev@...r.kernel.org>
Subject: Re: [TEST]  net/bind_wildcard flakiness

From: Jakub Kicinski <kuba@...nel.org>
Date: Fri, 16 May 2025 10:05:42 -0700
> Hi!
> 
> The bind_wildcard used to be rock solid, no flakes in last 1000 runs,
> but then around the time Eric posted his Rx optimizations we hit one
> flake, and now we hit another:

Thanks for the report.

The test only calls bind() so I guess it's not related to the RX
optimisation series.


> 
> List:
> https://netdev.bots.linux.dev/contest.html?test=bind-wildcard&executor=vmksft-net&pass=0
> 
> Outputs:
> https://netdev-3.bots.linux.dev/vmksft-net/results/122022/14-bind-wildcard/stdout
> https://netdev-3.bots.linux.dev/vmksft-net/results/123463/14-bind-wildcard/stdout
> 
> One does:
> 
> # 0.21 [+0.00] # bind_wildcard.c:768:reuseaddr:Expected ret (-1) == 0 (0)
> # 0.21 [+0.00] # reuseaddr: Test terminated by assertion
> # 0.21 [+0.00] #          FAIL  bind_wildcard.v6_any_only_v6_v4mapped_any.reuseaddr
> 
> The other:
> 
> # 0.25 [+0.00] # bind_wildcard.c:775:plain:Expected ret (-1) == 0 (0)
> # 0.25 [+0.00] # plain: Test terminated by assertion
> # 0.25 [+0.00] #          FAIL  bind_wildcard.v6_local_v6_v4mapped_any.plain

This is weird because both cases failed when bind() should succeed.

Will try to reproduce on my end.


> 
> 
> The only change on the infra side is that I increased disk IO cap,
> so the builds are now faster.