Displaying 5 results from an estimated 5 matches for "c38e39c378f4".
2019 Jun 13
0
memory leak in vhost_net_ioctl
...6/entry/common.c:301
> [<0000000029cc8ca7>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
>
> End of syzbot report.
>
> The function vhost_net_ubuf_alloc() appears in the two cases of dump info, for
> pid 7130 and 7117, suggesting that it is ubuf leak.
>
> Since commit c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
> the function vhost_net_flush() had been no longer releasing ubuf.
>
> Freeing the slab after the last flush in the release path fixes it.
>
>
> Fixes: c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_fl...
2019 Jun 13
2
memory leak in vhost_net_ioctl
...t;] do_syscall_64+0x76/0x1a0 arch/x86/entry/common.c:301
[<0000000029cc8ca7>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
End of syzbot report.
The function vhost_net_ubuf_alloc() appears in the two cases of dump info, for
pid 7130 and 7117, suggesting that it is ubuf leak.
Since commit c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
the function vhost_net_flush() had been no longer releasing ubuf.
Freeing the slab after the last flush in the release path fixes it.
Fixes: c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
Reported-by: Syz...
2019 Jun 13
2
memory leak in vhost_net_ioctl
...t;] do_syscall_64+0x76/0x1a0 arch/x86/entry/common.c:301
[<0000000029cc8ca7>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
End of syzbot report.
The function vhost_net_ubuf_alloc() appears in the two cases of dump info, for
pid 7130 and 7117, suggesting that it is ubuf leak.
Since commit c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
the function vhost_net_flush() had been no longer releasing ubuf.
Freeing the slab after the last flush in the release path fixes it.
Fixes: c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
Reported-by: Syz...
2019 Jun 14
2
memory leak in vhost_net_ioctl
...gcc (GCC) 9.0.0 20181231 (experimental)
> patch: https://syzkaller.appspot.com/x/patch.diff?x=11ff0de1a00000
>
And I want to try again the following tiny diff made based on the logic:
1_> vhost_net_ubuf_alloc() in the dump info suggests that it is ubuf leak.
2_> commit c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
makes vhost_net_flush() no longer release ubuf.
3_> in both reset_owner and release pathes, see vhost_net_reset_owner() and
vhost_net_release() please, vq is reset in wake of flush:
vhost_net_flush(n);
vhost_dev_sto...
2019 Jun 14
2
memory leak in vhost_net_ioctl
...gcc (GCC) 9.0.0 20181231 (experimental)
> patch: https://syzkaller.appspot.com/x/patch.diff?x=11ff0de1a00000
>
And I want to try again the following tiny diff made based on the logic:
1_> vhost_net_ubuf_alloc() in the dump info suggests that it is ubuf leak.
2_> commit c38e39c378f4 ("vhost-net: fix use-after-free in vhost_net_flush")
makes vhost_net_flush() no longer release ubuf.
3_> in both reset_owner and release pathes, see vhost_net_reset_owner() and
vhost_net_release() please, vq is reset in wake of flush:
vhost_net_flush(n);
vhost_dev_sto...