[Bug 283903] rtw88: possible skb leak
- In reply to: bugzilla-noreply_a_freebsd.org: "[Bug 283903] rtw88: possible skb leak"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sat, 11 Jan 2025 15:35:29 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=283903 --- Comment #7 from Guillaume Outters <guillaume-freebsd@outters.eu> --- Whoops, I got a new one! After an uptime of more than 2 days without any problem, I once again had two killed processes yesterday evening (while building GHC, the Haskell compiler), and then this morning while trying a big rsync, an "skb alloc failed" loop happened. Jan 8 16:01:44 Boot Jan 10 22:26:56 pasdfric kernel: pid 9508 (firefox), jid 0, uid 1001, was killed: failed to reclaim memory Jan 10 22:36:53 pasdfric kernel: pid 78425 (ghc), jid 0, uid 1001, was killed: failed to reclaim memory Jan 11 11:20:18 pasdfric kernel: rtw880: ERROR lkpi_80211_txq_tx_one: skb alloc failed Jan 11 11:20:19 pasdfric kernel: rtw880: ERROR lkpi_80211_txq_tx_one: skb alloc failed Jan 11 11:21:38 pasdfric kernel: rtw880: ERROR lkpi_80211_txq_tx_one: skb alloc failed […] At 14:00 my rsync still ran, but at about 10 KB/s. Here is the vmstat -m (with everything lkpi-related): Type Use Memory Req Size(s) lkpi80211 35 42112 39 128,256,1024,2048,32768 lkpikmalloc 7531 17555504 260476706 16,32,64,128,256,384,512,1024,2048,4096,8192,16384,32768,65536 lkpifw 12 384 24 32,128 lkpiskb 580877 2391855104 8175104 If I read well that makes *2 GB* for lkpiskb??? After a reboot, the rsync finished really quickly (can't estimate at which speed, but many MB/s) -- You are receiving this mail because: You are on the CC list for the bug.