On Thu, Jul 05, 2018 at 04:50:00PM +0800, Yan, Zheng wrote: > > > > On Jul 5, 2018, at 13:44, Luciano Coelho <luciano.coelho@xxxxxxxxx> wrote: > > > > On Thu, 2018-07-05 at 08:49 +0800, Yan, Zheng wrote: > >> I started to get following warning and unstable connection after > >> upgrading kernel to 4.17.3 (from 4.17.2). > >> > >> > >> iwlwifi 0000:04:00.0: loaded firmware version 38.c0e03d94.0 op_mode iwlmvm > >> iwlwifi 0000:04:00.0: Detected Intel(R) Dual Band Wireless AC 9260, REV=0x324 > >> audit: type=1130 audit(1529640588.009:69): pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-rfkill comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success' > >> iwlwifi 0000:04:00.0: base HW address: 74:e5:f9:a6:bb:52 > >> … > >> > >> > >> WARNING: CPU: 1 PID: 0 at drivers/net/wireless/intel/iwlwifi/mvm/tx.c:710 iwl_mvm_tx_tso_segment+0x28a/0x2a0 [iwlmvm] > > > > Strange, I don't see anything that could directly cause this issue: > > > > luca@redipa:~/public/linux$ git log v4.17.2..v4.17.3 --oneline -- net/ drivers/net/wireless/intel/iwlwifi/ > > 1d79bd39cd5b iwlwifi: fw: harden page loading code > > aba9ac38845c net: in virtio_net_hdr only add VLAN_HLEN to csum_start if payload holds vlan > > e8b9a16954b5 udp: fix rx queue len reported by diag and proc interface > > be3bb23cc0e4 socket: close race condition between sock_close() and sockfs_setattr() > > 6196f30e84ca tls: fix waitall behavior in tls_sw_recvmsg > > 7fd98de479ea tls: fix use-after-free in tls_push_record > > 1caf7914a7a5 tcp: verify the checksum of the first data segment in a new connection > > a316034b203f net/sched: act_simple: fix parsing of TCA_DEF_DATA > > 36ee02c5ff46 net: dsa: add error handling for pskb_trim_rcsum > > b788fc5f6977 ipv6: allow PMTU exceptions to local routes > > > > > > Did you happen to upgrade the firmware at the same time? How often does > > it happen? Can you try to bisect? > > Firmware wasn’t upgraded. I changed kernel back to 4.17.2, the issue no longer happens. Can you run 'git bisect' to track down the specific patch that causes the problem? Does 4.18-rc3 also have this problem? thanks, greg k-h
