On 2020-02-28 21:57, Todd C. Miller wrote:
This sounds like the loop in softdep_process_worklist() is never
exiting. It shouldn't run for more than a second, though.
FreeBSD breaks out of the loop if process_worklist_item() can't
make progress. You could try the following (untested) diff to see
if it changes the behavior.
After ~11h with your diff, the system was still up and running,
so i decided to reboot. It crashed while syncing disks.
syncing disks...uvm_fault(0xfffffd83a134d668, 0x20, 0, 1) -> e
kernel: page fault trap, code=0
Stopped at handle_workitem_freefile+0x2a: movq 0x20(%rax),%rcx
ddb{0}>
ddb{0}> trace
handle_workitem_freefile(fffffd8369e3be70) at handle_workitem_freefile+0x2a
process_worklist_item(ffff8000001b8800,40) at process_worklist_item+0x1c9
softdep_process_worklist(ffff8000001b8800) at softdep_process_worklist+0xed
softdep_flushworklist(ffff8000001b8800,ffff800022516218,ffff800022556d58) at so
ftdep_flushworklist+0xb8
ffs_sync(ffff8000001b8800,1,0,fffffd841f7c2480,ffff800022556d58) at
ffs_sync+0xdd
dounmount_leaf(ffff8000001b8800,80000,ffff800022556d58) at dounmount_leaf+0xaa
dounmount(ffff8000001b8800,80000,ffff800022556d58) at dounmount+0xfc
vfs_unmountall() at vfs_unmountall+0x8e
vfs_shutdown(ffff800022556d58) at vfs_shutdown+0x3b
boot(0) at boot+0x6c
reboot(0) at reboot+0x5c
sys_reboot(ffff800022556d58,ffff800022516470,ffff8000225164d0) at
sys_reboot+0x7e
syscall(ffff800022516540) at syscall+0x389
Xsyscall() at Xsyscall+0x128
end of kernel
end trace frame: 0x7f7ffffee610, count: -14
ddb{0}> ps
PID TID PPID UID S FLAGS WAIT COMMAND
*47072 417851 1 0 7 0x3 reboot
95768 197147 0 0 3 0x14200 bored smr
23039 211890 0 0 2 0x14200 zerothread
35233 64895 0 0 3 0x14200 aiodoned aiodoned
71347 316690 0 0 3 0x14200 syncer update
44681 504129 0 0 3 0x14200 cleaner cleaner
90819 193944 0 0 3 0x14200 reaper reaper
96855 230075 0 0 3 0x14200 pgdaemon pagedaemon
47917 12510 0 0 3 0x14200 bored crynlk
20365 68713 0 0 3 0x14200 bored crypto
21791 476177 0 0 3 0x14200 usbtsk usbtask
70170 396785 0 0 3 0x14200 usbatsk usbatsk
44347 23905 0 0 3 0x40014200 acpi0 acpi0
27912 30285 0 0 7 0x40014200 idle7
22700 306163 0 0 7 0x40014200 idle6
63705 100085 0 0 7 0x40014200 idle5
17377 411898 0 0 7 0x40014200 idle4
84659 510633 0 0 7 0x40014200 idle3
54830 118160 0 0 7 0x40014200 idle2
54818 330925 0 0 7 0x40014200 idle1
70767 292562 0 0 3 0x14200 bored sensors
44681 504129 0 0 3 0x14200 cleaner cleaner
90819 193944 0 0 3 0x14200 reaper reaper
96855 230075 0 0 3 0x14200 pgdaemon pagedaemon
47917 12510 0 0 3 0x14200 bored crynlk
20365 68713 0 0 3 0x14200 bored crypto
21791 476177 0 0 3 0x14200 usbtsk usbtask
70170 396785 0 0 3 0x14200 usbatsk usbatsk
44347 23905 0 0 3 0x40014200 acpi0 acpi0
27912 30285 0 0 7 0x40014200 idle7
22700 306163 0 0 7 0x40014200 idle6
63705 100085 0 0 7 0x40014200 idle5
17377 411898 0 0 7 0x40014200 idle4
84659 510633 0 0 7 0x40014200 idle3
54830 118160 0 0 7 0x40014200 idle2
54818 330925 0 0 7 0x40014200 idle1
70767 292562 0 0 3 0x14200 bored sensors
59264 346295 0 0 3 0x14200 bored softnet
10461 133156 0 0 3 0x14200 bored systqmp
1863 323920 0 0 3 0x14200 bored systq
11411 181100 0 0 2 0x40014200 softclock
3128 136983 0 0 3 0x40014200 idle0
1 52990 0 0 3 0x82 wait init
0 0 -1 0 3 0x10200 scheduler swapper
ddb{0}> show registers
rdi 0xfffffd8369e3be70
rsi 0x40
rbp 0xffff800022515f50
rbx 0x1
rdx 0xfe00000007ff1e3a
rcx 0x202
rax 0
r8 0
r9 0xffffffffffffffff
r10 0x208d459308eb8ce2
r11 0x572f46b22ffb6c93
r12 0x40
r13 0x1
r14 0xfffffd8369e3be70
r15 0xffff8000001b8800
rip 0xffffffff812fb35a handle_workitem_freefile+0x2a
cs 0x8
rflags 0x10282 __ALIGN_SIZE+0xf282
rsp 0xffff800022515d70
ss 0x10
handle_workitem_freefile+0x2a: movq 0x20(%rax),%rcx
-Mark
- todd
Index: /sys/ufs/ffs/ffs_softdep.c
===================================================================
RCS file: /cvs/src/sys/ufs/ffs/ffs_softdep.c,v
retrieving revision 1.148
diff -u -p -u -r1.148 ffs_softdep.c
--- /sys/ufs/ffs/ffs_softdep.c 4 Feb 2020 04:09:11 -0000 1.148
+++ /sys/ufs/ffs/ffs_softdep.c 28 Feb 2020 20:42:39 -0000
@@ -591,7 +591,7 @@ int
softdep_process_worklist(struct mount *matchmnt)
{
struct proc *p = CURPROC;
- int matchcnt, loopcount;
+ int cnt, matchcnt, loopcount;
struct timeval starttime;
/*
@@ -639,7 +639,9 @@ softdep_process_worklist(struct mount *m
loopcount = 1;
getmicrouptime(&starttime);
while (num_on_worklist > 0) {
- matchcnt += process_worklist_item(matchmnt, LK_NOWAIT);
+ if ((cnt = process_worklist_item(matchmnt, LK_NOWAIT)) == 0)
+ break;
+ matchcnt += cnt;
/*
* If a umount operation wants to run the worklist
--
Mark Patruck ( mark at wrapped.cx )
GPG key 0xF2865E51 / 187F F6D3 EE04 1DCE 1C74 F644 0D3C F66F F286 5E51
https://www.wrapped.cx
