On Tue, May 10, 2022 at 6:16 AM Steve Sakoman via
lists.openembedded.org <steve=sakoman....@lists.openembedded.org>
wrote:
>
> On Tue, May 10, 2022 at 5:07 AM Ranjitsinh Rathod
> <ranjitsinhrathod1...@gmail.com> wrote:
>
> > It seems something is wrong with this patch and also in the patching tool
> > within Yocto as well.
> >
> > When I apply this patch, it is getting applied cleanly, without any
> > warnings and/or patch-fuzz, which means the source code should reflect as
> > per the patch.
> > But when I go to the source code on the path
> > "build/tmp/work/<architecture>/libinput/1.15.2-r0/libinput-1.15.2" and see
> > the file "src/evdev.c", it seems below two Hunk's are not getting reflected.
>
> Yikes! This is quite concerning.
>
> I can confirm that these two hunks are not being applied. I also
> checked the hunks before and after these two and those seem to be
> applied correctly.
>
> Looking at your CVE patch, it does seem that there is an extra newline
> before these two hunks. I tried removing the newline, but the 2 hunks
> are still not applied.
My mistake above -- removing the extra newline in your patch does fix
the issue. I will correct your patch by removing the extra newline, no
need to resubmit.
Whether this is something that patching tools should catch is another matter!
Steve
> I then tried deliberately malforming the first of the hunks to see if
> it was just being ignored for some reason. However, as expected the
> patch process failed at the malformed hunk, so this part of the patch
> is definitely being parsed.
>
> Will keep looking into this, but would appreciate help from any
> experts in the patching process!
>
> Steve
>
> >
> > @@ -2469,7 +2478,7 @@ evdev_device_get_output(struct evdev_device *device)
> > const char *
> > evdev_device_get_sysname(struct evdev_device *device)
> > {
> > - return udev_device_get_sysname(device->udev_device);
> > + return device->sysname;
> > }
> >
> > const char *
> > @@ -3066,6 +3075,8 @@ evdev_device_destroy(struct evdev_device *device)
> > if (device->base.group)
> > libinput_device_group_unref(device->base.group);
> >
> > + free(device->log_prefix_name);
> > + free(device->sysname);
> > free(device->output_name);
> > filter_destroy(device->pointer.filter);
> > libinput_timer_destroy(&device->scroll.timer);
> >
> > Can you please check this from your end too? If this is the case, there is
> > a serious issue where patching is failing silently which needs attention.
> >
> > As I have seen the patch in the testing branch
> > "http://cgit.openembedded.org/openembedded-core-contrib/commit/?h=stable/dunfell-nut&id=6a31bfaee4d715f31e69cea1eed05484d9d7a999";
> > I considered to post this email as it seems you also did not get any
> > warnings/error during the patching and even compilation.
> >
> > Thanks,
> > Ranjitsinh Rathod
> >
> >
> >
>
>
>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#165465):
https://lists.openembedded.org/g/openembedded-core/message/165465
Mute This Topic: https://lists.openembedded.org/mt/90983830/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
