* syzkaller config
@ 2021-05-13 15:41 Christoph Paasch
2021-05-13 22:05 ` Mat Martineau
0 siblings, 1 reply; 2+ messages in thread
From: Christoph Paasch @ 2021-05-13 15:41 UTC (permalink / raw)
To: mptcp
Hello,
this is my syzkaller config:
{
"target": "linux/amd64",
"http": "127.0.0.1:56741",
"workdir": "/home/cpaasch/gopath/src/github.com/google/syzkaller/workdir_kasan_net",
"kernel_obj": "/mnt/tmp/build/",
"kernel_src": "/mnt/tmp/mptcp_syzkaller/",
"kernel_build_src": "/mnt/tmp/mptcp_syzkaller/",
"image": "/home/cpaasch/syzkaller/tools/stretch.img",
"sshkey": "/home/cpaasch/syzkaller/tools/stretch.id_rsa",
"syzkaller": "/home/cpaasch/gopath/src/github.com/google/syzkaller",
"disable_syscalls": ["perf_event_open", "syz_mount_image", "syz_read_part_table", "openat$ttyprintk", "mount", "mkdir", "openat$ptmx", "mq_open", "fsetxattr", "rt_tgsigqueueinfo", "ioctl$VT_RESIZE", "ioctl$TIOCVHANGUP", "get_robust_list", "openat$nullb", "ioctl$SCSI_IOCTL_SEND_COMMAND", "lremovexattr", "mknod$loop", "write$binfmt_script", "syz_open_dev$sg", "write$nbd", "prlimit64", "write$P9_RRENAMEAT", "fcntl$addseals", "finit_module", "ioctl$KDSETMODE", "write$FUSE_NOTIFY_STORE", "ioctl$TIOCL_SETVESABLANK", "fsmount", "socket$vsock_stream", "socketpair$unix", "socket$nl_audit", "connect$unix", "bind$unix"],
"procs": 8,
"type": "qemu",
"vm": {
"count": 24,
"kernel": "/mnt/tmp/build/arch/x86/boot/bzImage",
"cpu": 2,
"mem": 4096
}
}
The way I generate this list of disabled syscals is simply over time, when I
saw that syzkaller was spending all of its efforts in those syscalls, I
added it to the disable-list. That way now it is spending more time in
MPTCP/networking related code.
Christoph
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: syzkaller config
2021-05-13 15:41 syzkaller config Christoph Paasch
@ 2021-05-13 22:05 ` Mat Martineau
0 siblings, 0 replies; 2+ messages in thread
From: Mat Martineau @ 2021-05-13 22:05 UTC (permalink / raw)
To: Christoph Paasch; +Cc: mptcp
On Thu, 13 May 2021, Christoph Paasch wrote:
> Hello,
>
> this is my syzkaller config:
>
> {
> "target": "linux/amd64",
> "http": "127.0.0.1:56741",
> "workdir": "/home/cpaasch/gopath/src/github.com/google/syzkaller/workdir_kasan_net",
> "kernel_obj": "/mnt/tmp/build/",
> "kernel_src": "/mnt/tmp/mptcp_syzkaller/",
> "kernel_build_src": "/mnt/tmp/mptcp_syzkaller/",
> "image": "/home/cpaasch/syzkaller/tools/stretch.img",
> "sshkey": "/home/cpaasch/syzkaller/tools/stretch.id_rsa",
> "syzkaller": "/home/cpaasch/gopath/src/github.com/google/syzkaller",
> "disable_syscalls": ["perf_event_open", "syz_mount_image", "syz_read_part_table", "openat$ttyprintk", "mount", "mkdir", "openat$ptmx", "mq_open", "fsetxattr", "rt_tgsigqueueinfo", "ioctl$VT_RESIZE", "ioctl$TIOCVHANGUP", "get_robust_list", "openat$nullb", "ioctl$SCSI_IOCTL_SEND_COMMAND", "lremovexattr", "mknod$loop", "write$binfmt_script", "syz_open_dev$sg", "write$nbd", "prlimit64", "write$P9_RRENAMEAT", "fcntl$addseals", "finit_module", "ioctl$KDSETMODE", "write$FUSE_NOTIFY_STORE", "ioctl$TIOCL_SETVESABLANK", "fsmount", "socket$vsock_stream", "socketpair$unix", "socket$nl_audit", "connect$unix", "bind$unix"],
> "procs": 8,
> "type": "qemu",
> "vm": {
> "count": 24,
You're using a much bigger system than me :)
> "kernel": "/mnt/tmp/build/arch/x86/boot/bzImage",
> "cpu": 2,
> "mem": 4096
> }
> }
>
>
> The way I generate this list of disabled syscals is simply over time, when I
> saw that syzkaller was spending all of its efforts in those syscalls, I
> added it to the disable-list. That way now it is spending more time in
> MPTCP/networking related code.
I based the sample config in
https://github.com/multipath-tcp/mptcp_net-next/wiki/Testing#syzkaller on
this (with a couple of adjustments and some anonymization). Thanks for
sharing.
--
Mat Martineau
Intel
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2021-05-13 22:05 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-13 15:41 syzkaller config Christoph Paasch
2021-05-13 22:05 ` Mat Martineau
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).