qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v2 0/6] esp: fix asserts/segfaults discovered by fuzzer


From: Mark Cave-Ayland
Subject: Re: [PATCH v2 0/6] esp: fix asserts/segfaults discovered by fuzzer
Date: Tue, 30 Mar 2021 08:34:51 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.9.0

On 18/03/2021 18:13, Paolo Bonzini wrote:

On 18/03/21 00:02, Mark Cave-Ayland wrote:
Recently there have been a number of issues raised on Launchpad as a result of
fuzzing the am53c974 (ESP) device. I spent some time over the past couple of
days checking to see if anything had improved since my last patchset: from
what I can tell the issues are still present, but the cmdfifo related failures
now assert rather than corrupting memory.

This patchset applied to master passes my local tests using the qtest fuzz test
cases added by Alexander for the following Launchpad bugs:

   https://bugs.launchpad.net/qemu/+bug/1919035
   https://bugs.launchpad.net/qemu/+bug/1919036
   https://bugs.launchpad.net/qemu/+bug/1910723
   https://bugs.launchpad.net/qemu/+bug/1909247
I'm posting this now just before soft freeze since I see that some of the issues
have recently been allocated CVEs and so it could be argued that even though
they have existed for some time, it is worth fixing them for 6.0.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>

v2:
- Add Alexander's R-B tag for patch 2 and Phil's R-B for patch 3
- Add patch 4 for additional testcase provided in Alexander's patch 1 comment
- Move current_req NULL checks forward in DMA functions (fixes ASAN bug reported
   at https://bugs.launchpad.net/qemu/+bug/1909247/comments/6) in patch 3
- Add qtest for am53c974 containing a basic set of regression tests using the
   automatic test cases generated by the fuzzer as requested by Paolo


Mark Cave-Ayland (6):
   esp: don't underflow cmdfifo if no message out/command data is present
   esp: don't overflow cmdfifo if TC is larger than the cmdfifo size
   esp: ensure cmdfifo is not empty and current_dev is non-NULL
   esp: don't underflow fifo when writing to the device
   esp: always check current_req is not NULL before use in DMA callbacks
   tests/qtest: add tests for am53c974 device

  hw/scsi/esp.c               |  73 +++++++++++++--------
  tests/qtest/am53c974-test.c | 122 ++++++++++++++++++++++++++++++++++++
  tests/qtest/meson.build     |   1 +
  3 files changed, 171 insertions(+), 25 deletions(-)
  create mode 100644 tests/qtest/am53c974-test.c


Queued, thanks.

Paolo

Hi Paolo,

I had a quick look at Alex's updated test cases and most of them are based on an incorrect assumption I made around the behaviour of fifo8_pop_buf(). Can you drop these for now, and I will submit a v3 shortly once I've given it a full run through my test images?


ATB,

Mark.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]