CRIU hang on dump or restore

[biozit]

I am trying to dump or restore any process (even a process that no exists) using sudo or root access:

root@test:~#  criu dump -t 1211114 -vvvvv
(00.000005) Version: 3.17.1 (gitid 0)
(00.000026) Running on osdftest.t2.ucsd.edu Linux 5.4.0-135-generic #152-Ubuntu SMP Wed Nov 23 20:19:22 UTC 2022 x86_64
(00.000041) File /run/criu.kdat does not exist
(00.000065) sockets: Probing sock diag modules
(00.000128) sockets: Done probing
(00.027251) Pagemap is fully functional
(00.027305) Found anon-shmem device at 1
(00.027339) Hugetlb size 2 Mb is supported but cannot get dev's number
(00.027358) Hugetlb size 1024 Mb is supported but cannot get dev's number
(00.027376) Reset 46767's dirty tracking
(00.027461)  ... done
(00.027502) Dirty track supported on kernel
(00.027589) Found task size of 7ffffffff000

kernel:

root@test:~# uname -a
Linux  5.4.0-135-generic #152-Ubuntu SMP Wed Nov 23 20:19:22 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

No error msg. It only hangs.

[avagin]

Run it under strace and attach logs here:

$ strace -s 1024 -f -o strace.log criu dump -t XXX -vvvv 

[biozit]

strace.log

[avagin]

Could you try out with this patch?

diff --git a/criu/kerndat.c b/criu/kerndat.c
index 5b567e79f..a2d4e2a89 100644
--- a/criu/kerndat.c
+++ b/criu/kerndat.c
@@ -599,7 +599,7 @@ static int kerndat_loginuid(void)
 static int kerndat_iptables_has_xtlocks(void)
 {
        int fd;
-       char *argv[4] = { "sh", "-c", "iptables -w -L", NULL };
+       char *argv[4] = { "sh", "-c", "iptables -n -w -L", NULL };
 
        fd = open("/dev/null", O_RDWR);
        if (fd < 0) {

[biozit]

Hi, now it is working! Thank you!

[github-actions]

A friendly reminder that this issue had no activity for 30 days.

[biozit]

Please, could I help with this?

[github-actions]

A friendly reminder that this issue had no activity for 30 days.

[hanwen-flow]

I am seeing this too, but only on the first checkpoint after boot.

This is with

root@ip-172-31-7-139:~# criu --version
Version: 3.17.1
root@ip-172-31-7-139:~# uname -a
Linux ip-172-31-7-139 6.1.0-31-cloud-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.128-1 (2025-02-07) x86_64 GNU/Linux

the process isn't hung, it continues after 40 seconds,

(00.030876) Found anon-shmem device at 1
(00.031605) Hugetlb size 2 Mb is supported but cannot get dev's number
(00.031614) Reset 834's dirty tracking
(00.031662)  ... done
(00.031687) Dirty track supported on kernel
(00.033115) Found task size of 7ffffffff000
(40.088859) Restoring netdev veth idx 10
(40.090884) Dumping netns links
(40.090911)     LD: Got link 1, type 772
(40.090914)     LD: Got link 10, type 1
(40.091813) vdso: Parsing at 7ffd545e8000 7ffd545ea000
(40.091821) vdso: PT_LOAD p_vaddr: 0

[hanwen-flow]

This reproduces with the latest CRIU version and the fix to add '-n' to the iptables command worked.

[hanwen-flow]

@adrianreber - is there anything against submitting the patch in #2032 (comment) ?