On Fri, Jul 17, 2020 at 10:07 AM Paul Menzel <pmenzel@xxxxxxxxxxxxx> wrote: > > Dear Linux folks, > > > On Debian Sid/unstable with Linux 5.7.6, `ls` hangs sometimes when > accessing a directory with an SSHFS mount. Linux logs the messages below. Several solutions: - kill `pidof sshfs` - umount -f $MOUNTPOINT - echo 1 > /sys/fs/fuse/connections/$DEVNUM/abort > > ``` > [105591.121285] INFO: task ls:21242 blocked for more than 120 seconds. > [105591.121293] Not tainted 5.7.0-1-amd64 #1 Debian 5.7.6-1 > [105591.121295] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" > disables this message. > [105591.121298] ls D 0 21242 778 0x00004004 > [105591.121304] Call Trace: > [105591.121319] __schedule+0x2da/0x770 > [105591.121326] schedule+0x4a/0xb0 > [105591.121339] request_wait_answer+0x122/0x210 [fuse] > [105591.121349] ? finish_wait+0x80/0x80 > [105591.121357] fuse_simple_request+0x198/0x290 [fuse] > [105591.121366] fuse_do_getattr+0xcf/0x2c0 [fuse] > [105591.121376] vfs_statx+0x96/0xe0 > [105591.121382] __do_sys_statx+0x3b/0x80 > [105591.121391] do_syscall_64+0x52/0x180 > [105591.121396] entry_SYSCALL_64_after_hwframe+0x44/0xa9 > [105591.121400] RIP: 0033:0x7f948c1c972a > [105591.121410] Code: Bad RIP value. > [105591.121412] RSP: 002b:00007ffd94582dd8 EFLAGS: 00000246 ORIG_RAX: > 000000000000014c > [105591.121416] RAX: ffffffffffffffda RBX: 0000556c6a957cc8 RCX: > 00007f948c1c972a > [105591.121417] RDX: 0000000000000100 RSI: 00007ffd94582f10 RDI: > 00000000ffffff9c > [105591.121419] RBP: 000000000000025e R08: 00007ffd94582de0 R09: > 000000006a95c700 > [105591.121421] R10: 000000000000025e R11: 0000000000000246 R12: > 0000556c6a95c763 > [105591.121423] R13: 0000000000000003 R14: 00007ffd94582f10 R15: > 0000556c6a957cc8 > ``` > > The `ls` process cannot be killed. The SSHFS issue *Fuse sshfs blocks > standby (Visual Studio Code?)* from 2018 already reported this for Linux > 4.17, and the SSHFS developers asked to report this to the Linux kernel. This is a very old and fundamental issue. Theoretical solution for killing the stuck process exists, but it's not trivial and since the above mentioned workarounds work well in all cases it's not high priority right now. Thanks, Miklos
