Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption
From: Jakub Kicinski <kuba@kernel.org>
Date: 2026-05-21 15:56:03
Possibly related (same subject, not in this thread)
- 2026-05-23 · Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption · Jamal Hadi Salim <jhs@mojatatu.com>
- 2026-05-23 · Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption · Jakub Kicinski <kuba@kernel.org>
- 2026-05-23 · Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption · Jamal Hadi Salim <jhs@mojatatu.com>
- 2026-05-23 · Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption · Jamal Hadi Salim <jhs@mojatatu.com>
- 2026-05-23 · Re: [PATCH net] net/sched: fix pedit partial COW leading to page cache corruption · Jakub Kicinski <kuba@kernel.org>
On Thu, 21 May 2026 17:51:25 +0200 Davide Caratti wrote:
quoted
Do you know of anyone modifying payloads with pedit?yes: MPTCP kselftest does that on purpose, to mimic a middlebox and trigger DSS checksum failures [1]
As I said in another reply (looks like the ML is having a significant lag this morning :() - we would still support it, the shared frag tracking is just an optimization to avoid a copy. And maintaining that optimization for an occasional CI test is not the right eng tradeoff.