Changes installed in d22d7d6be8c

Changes installed in d22d7d6be8c

[Po Lu]

I see the following code in nsterm.m:

d22d7d6be8c (Alan Third            2021-09-27  7934)   NSRect dstRect = NSMakeRect (dest.x, dest.y, NSWidth (srcRect),
d22d7d6be8c (Alan Third            2021-09-27  7935)                                NSHeight (srcRect));
d22d7d6be8c (Alan Third            2021-09-27  7936)   NSRect frame = [self frame];
d22d7d6be8c (Alan Third            2021-09-27  7937) 
d22d7d6be8c (Alan Third            2021-09-27  7938)   /* TODO: This check is an attempt to debug a rare graphical glitch
d22d7d6be8c (Alan Third            2021-09-27  7939)      on macOS and should be removed before the Emacs 28 release.  */
d22d7d6be8c (Alan Third            2021-09-27  7940)   if (!NSContainsRect (frame, srcRect)
d22d7d6be8c (Alan Third            2021-09-27  7941)       || !NSContainsRect (frame, dstRect))
d22d7d6be8c (Alan Third            2021-09-27  7942)     {
d22d7d6be8c (Alan Third            2021-09-27  7943)       NSLog (@"[EmacsView copyRect:to:] Attempting to copy to or "
d22d7d6be8c (Alan Third            2021-09-27  7944)              "from an area outside the graphics buffer.");
d22d7d6be8c (Alan Third            2021-09-27  7945)       NSLog (@"  Frame: (%f, %f) %f×%f",
d22d7d6be8c (Alan Third            2021-09-27  7946)              NSMinX (frame), NSMinY (frame),
d22d7d6be8c (Alan Third            2021-09-27  7947)              NSWidth (frame), NSHeight (frame));
d22d7d6be8c (Alan Third            2021-09-27  7948)       NSLog (@"  Source: (%f, %f) %f×%f",
d22d7d6be8c (Alan Third            2021-09-27  7949)              NSMinX (srcRect), NSMinY (srcRect),
d22d7d6be8c (Alan Third            2021-09-27  7950)              NSWidth (srcRect), NSHeight (srcRect));
d22d7d6be8c (Alan Third            2021-09-27  7951)       NSLog (@"  Destination: (%f, %f) %f×%f",
d22d7d6be8c (Alan Third            2021-09-27  7952)              NSMinX (dstRect), NSMinY (dstRect),
d22d7d6be8c (Alan Third            2021-09-27  7953)              NSWidth (dstRect), NSHeight (dstRect));
d22d7d6be8c (Alan Third            2021-09-27  7954)     }

Could someone explain exactly how the graphical glitch here manifests?

It's also getting close to the Emacs 28 release, so maybe this should be
removed there.

Thanks in advance.

Re: Changes installed in d22d7d6be8c

[Alan Third]

On Sat, Jan 29, 2022 at 07:30:29PM +0800, Po Lu wrote:
> 
> I see the following code in nsterm.m:
> 
> d22d7d6be8c (Alan Third            2021-09-27  7934)   NSRect dstRect = NSMakeRect (dest.x, dest.y, NSWidth (srcRect),
> d22d7d6be8c (Alan Third            2021-09-27  7935)                                NSHeight (srcRect));
> d22d7d6be8c (Alan Third            2021-09-27  7936)   NSRect frame = [self frame];
> d22d7d6be8c (Alan Third            2021-09-27  7937) 
> d22d7d6be8c (Alan Third            2021-09-27  7938)   /* TODO: This check is an attempt to debug a rare graphical glitch
> d22d7d6be8c (Alan Third            2021-09-27  7939)      on macOS and should be removed before the Emacs 28 release.  */
> d22d7d6be8c (Alan Third            2021-09-27  7940)   if (!NSContainsRect (frame, srcRect)
> d22d7d6be8c (Alan Third            2021-09-27  7941)       || !NSContainsRect (frame, dstRect))
> d22d7d6be8c (Alan Third            2021-09-27  7942)     {
> d22d7d6be8c (Alan Third            2021-09-27  7943)       NSLog (@"[EmacsView copyRect:to:] Attempting to copy to or "
> d22d7d6be8c (Alan Third            2021-09-27  7944)              "from an area outside the graphics buffer.");
> d22d7d6be8c (Alan Third            2021-09-27  7945)       NSLog (@"  Frame: (%f, %f) %f×%f",
> d22d7d6be8c (Alan Third            2021-09-27  7946)              NSMinX (frame), NSMinY (frame),
> d22d7d6be8c (Alan Third            2021-09-27  7947)              NSWidth (frame), NSHeight (frame));
> d22d7d6be8c (Alan Third            2021-09-27  7948)       NSLog (@"  Source: (%f, %f) %f×%f",
> d22d7d6be8c (Alan Third            2021-09-27  7949)              NSMinX (srcRect), NSMinY (srcRect),
> d22d7d6be8c (Alan Third            2021-09-27  7950)              NSWidth (srcRect), NSHeight (srcRect));
> d22d7d6be8c (Alan Third            2021-09-27  7951)       NSLog (@"  Destination: (%f, %f) %f×%f",
> d22d7d6be8c (Alan Third            2021-09-27  7952)              NSMinX (dstRect), NSMinY (dstRect),
> d22d7d6be8c (Alan Third            2021-09-27  7953)              NSWidth (dstRect), NSHeight (dstRect));
> d22d7d6be8c (Alan Third            2021-09-27  7954)     }
> 
> Could someone explain exactly how the graphical glitch here manifests?
> 
> It's also getting close to the Emacs 28 release, so maybe this should be
> removed there.

I thought this was already removed, but when I checked my local
emacs-28 branch the commit was still unpushed, so I've just done that.

Thanks.
-- 
Alan Third

Re: Changes installed in d22d7d6be8c

[Po Lu]

Alan Third <alan@idiocy.org> writes:

> I thought this was already removed, but when I checked my local
> emacs-28 branch the commit was still unpushed, so I've just done that.

Thanks.  I presume that means whatever problem it was supposed to debug
has been solved?

Re: Changes installed in d22d7d6be8c

[Alan Third]

On Sat, Jan 29, 2022 at 08:33:18PM +0800, Po Lu wrote:
> Alan Third <alan@idiocy.org> writes:
> 
> > I thought this was already removed, but when I checked my local
> > emacs-28 branch the commit was still unpushed, so I've just done that.
> 
> Thanks.  I presume that means whatever problem it was supposed to debug
> has been solved?

I think so, it was quite a while ago now and I don't have the emails
in my inbox any more, but we definitely ruled out the possibility that
it was copyRect going past the frame boundaries as nobody ever
reported seeing the log message.

-- 
Alan Third