bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Drew Adams]

emacs -Q

M-x load-library mouse.el
M-x debug-on-entry mouse-drag-region

Confirm that clicking mouse-1 opens the debugger for
`mouse-drag-region'.

M-: (cancel-debug-on-error)   ; (or `M-x cancel-debug-on-error')

Click mouse-1.  The debugger still opens for `mouse-drag-region'.
Debugging was not cancelled for all functions, as it should have been.


In GNU Emacs 24.4.50.1 (i686-pc-mingw32)
 of 2014-05-17 on ODIEONE
Bzr revision: 117119 eggert@cs.ucla.edu-20140517081131-ugu7ociaoec2xk7y
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --prefix=/c/Devel/emacs/snapshot/trunk
 --enable-checking=yes,glyphs 'CFLAGS=-O0 -g3'
 LDFLAGS=-Lc:/Devel/emacs/lib 'CPPFLAGS=-DGC_MCHECK=1
 -Ic:/Devel/emacs/include''

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Michael Heerdegen]

Drew Adams <drew.adams@oracle.com> writes:

> emacs -Q
>
> M-x load-library mouse.el
> M-x debug-on-entry mouse-drag-region
>
> Confirm that clicking mouse-1 opens the debugger for
> `mouse-drag-region'.
>
> M-: (cancel-debug-on-error)   ; (or `M-x cancel-debug-on-error')

Typo: you mean and used `cancel-debug-on-entry'.  But I can reproduce
it.

Here, `debug--function-list' returns nil after the `debug-on-entry' use,
so I guess it's related to nadvice (the according `advice--member-p'
test fails).

Michael.

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Drew Adams]

> > M-: (cancel-debug-on-error)   ; (or `M-x cancel-debug-on-error')
> 
> Typo: you mean and used `cancel-debug-on-entry'.

Yes, sorry.

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Stefan Monnier]

> Debugging was not cancelled for all functions, as it should have been.

Thanks.  I installed a fix for it in the emacs-24 branch.


        Stefan

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Drew Adams]

> > Debugging was not cancelled for all functions, as it should have been.
> 
> Thanks.  I installed a fix for it in the emacs-24 branch.

But not in the trunk, perhaps?  Should we expect that it is fixed in
trunk also?

In any case, the regression is *still present* in this build, which
was created two days after your fix was installed in the emacs-24
branch.

If the fix was not applied to trunk, can you please do so?  If it was,
then please reopen the bug, as this is not fixed.

emacs -Q
M-x debug-on-entry forward-char
C-f
M-x cancel-debug-on-entry RET
C-f

Still enters the debugger.

FWIW, this regression is really quite annoying.  It pretty much
requires that one restart Emacs and get back into the proper setup
to continue.  Where one might typically toggle debugging of a
function multiple times while working on it, this effectively
puts a stop to that.

In GNU Emacs 24.4.50.1 (i686-pc-mingw32)
 of 2014-05-25 on ODIEONE
Bzr revision: 117153 tsdh@gnu.org-20140525174054-vzeh4zeg00a1ley8
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --prefix=/c/Devel/emacs/snapshot/trunk
 --enable-checking=yes,glyphs 'CFLAGS=-O0 -g3'
 LDFLAGS=-Lc:/Devel/emacs/lib 'CPPFLAGS=-DGC_MCHECK=1
 -Ic:/Devel/emacs/include''

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Michael Heerdegen]

Hi Drew,

> > Thanks.  I installed a fix for it in the emacs-24 branch.
>
> But not in the trunk, perhaps?

Indeed.  AFAICT, the standard procedure for fixes in the release branch
is not to install in both branches in parallel, but, in our case, to
install only in emacs-24.  Changes are merged back into trunk regularly
(once a week?).  I guess you just have to wait some time.

> Should we expect that it is fixed in trunk also?

I think so.

Michael.

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Drew Adams]

> > > I installed a fix for it in the emacs-24 branch.
> > But not in the trunk, perhaps?
> 
> Indeed.  AFAICT, the standard procedure for fixes in the release
> branch is not to install in both branches in parallel, but, in our
> case, to install only in emacs-24.

> Changes are merged back into trunk regularly (once a week?).  I guess
> you just have to wait some time.
>
> > Should we expect that it is fixed in trunk also?
> 
> I think so.

I see.  Thx.  I never noticed such a policy before, for some reason.

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Stephen Berman]

On Tue, 27 May 2014 09:47:21 -0700 (PDT) Drew Adams <drew.adams@oracle.com> wrote:

>> > > I installed a fix for it in the emacs-24 branch.
>> > But not in the trunk, perhaps?
>> 
>> Indeed.  AFAICT, the standard procedure for fixes in the release
>> branch is not to install in both branches in parallel, but, in our
>> case, to install only in emacs-24.
>
>> Changes are merged back into trunk regularly (once a week?).  I guess
>> you just have to wait some time.
>>
>> > Should we expect that it is fixed in trunk also?
>> 
>> I think so.
>
> I see.  Thx.  I never noticed such a policy before, for some reason.

Here's a ref (the first lines):

http://bzr.savannah.gnu.org/lh/emacs/trunk/annotate/head:/admin/notes/bzr

Steve Berman

bug#17531: 24.4.50; REGRESSION: `cancel-debug-on-error' no longer works

[Stefan Monnier]

>> Thanks.  I installed a fix for it in the emacs-24 branch.
> But not in the trunk, perhaps?  Should we expect that it is fixed in
> trunk also?

For those who haven't followed Emacs's development practices over the
last couple years, here is how it works: the release branch (currently
called emacs-23, previously called emacs-23) gets merged manually every
once in a while (say once a week) back into trunk.


        Stefan