Re: [RFC PATCH] qemu-keymap: properly check return from xkb_keymap_mod_get_index

[Alex Bennée]

Peter Maydell <peter.maydell@linaro.org> writes:

> On Tue, 20 Jun 2023 at 16:04, Alex Bennée <alex.bennee@linaro.org> wrote:
>>
>> We can return XKB_MOD_INVALID which rightly gets flagged by sanitisers
>> as an overly wide shift attempt.
>>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> ---
>>  qemu-keymap.c | 24 ++++++++++++++++--------
>>  1 file changed, 16 insertions(+), 8 deletions(-)
>
> Looking at the code that works with the mask values
> we are getting here, I think this ought to work
> (if there's no AltGr modifier then the 0 mask means
> the key state will be the same in normal and with the
> altgr mask supplied, which we already check for).
> Did you eyeball the output, though?
>
> Also, which keymap runs into this? Is it every keymap
> for some new version of the xkb data (which would imply
> that the problem is that the AltGr modifier has changed
> name), or is it only one specific keymap that happens
> to have no AltGr key?

ar maybe? it only got flagged in clang-system once fedora was updated (I
assume with better sanitizers):

  [2773/3696] Generating pc-bios/keymaps/ar with a custom command
  FAILED: pc-bios/keymaps/ar 
  /builds/stsquad/qemu/build/qemu-keymap -f pc-bios/keymaps/ar -l ar
  ../qemu-keymap.c:223:16: runtime error: shift exponent 4294967295 is too 
large for 32-bit type 'int'
  SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior 
../qemu-keymap.c:223:16 in 
  [2774/3696] Generating pc-bios/edk2-x86_64-code.fd with a custom command 
(wrapped by meson to capture output)
  [2775/3696] Generating pc-bios/edk2-x86_64-secure-code.fd with a custom 
command (wrapped by meson to capture output)
  ninja: build stopped: subcommand failed.
  make: *** [Makefile:153: run-ninja] Error 1

https://gitlab.com/stsquad/qemu/-/jobs/4500683186#L3957

>
> thanks
> -- PMM


-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro