Re: freeze for 1.0.49

Juho Snellman <jsnell <at> iki.fi> writes:

> I'll be releasing 1.0.49 next weekend. Please let sbcl-devel know of any
> regressions.

I built the Quicklisp world with the latest checkout and everything
worked fine on Linux/amd64.

Zach

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

Re: freeze for 1.0.49

Nikodemus Siivola <nikodemus <at> random-state.net> writes:

> On 31 May 2011 02:30, Heka Treep <zena.treep <at> gmail.com> wrote:
>
>> Please accept these cosmetic changes I did before:
>> http://groups.google.com/group/sbcl-devel/browse_thread/thread/4e2dc3b372ef2888
>> (few patches in the last post).
>
> I have these in my queue for 1.0.49.early. Freeze period is pretty
> much reserved for fixing regressions.

Also in queue for 1.0.49.early is another round of SunOS test
cleanups -- a bunch of tests can now be re-enabled cleanly.

I've been thinking about skipped tests -- right now, if a test isn't
relevant on a platform, or crashes (as opposed to failing), we skip it,
but this doesn't get counted anywhere.

With the changes I have ready to go in, for instance, SunOS/x86 gives:

     Finished running tests.
     Status:
      Expected failure:    packages.impure.lisp / USE-PACKAGE-CONFLICT-SET
      Expected failure:    packages.impure.lisp / IMPORT-SINGLE-CONFLICT
     ok
     //apparent success (reached end of run-tests.sh normally)
     Wed Jun  1 11:39:23 EDT 2011

but in truth, all thread-related tests have been skipped (since this is
a non-threaded build), and a few other tests, such as TRACE-RECURSIVE

Re: freeze for 1.0.49

On Wed, Jun 01, 2011 at 01:47:18PM -0400, Jim Wise wrote:
> I'd like to change this to something like:
> 
>      Finished running tests.
>      Status:
>       Expected failure:    packages.impure.lisp / USE-PACKAGE-CONFLICT-SET
>       Expected failure:    packages.impure.lisp / IMPORT-SINGLE-CONFLICT
>       Skipped (broken):    debug.impure.lisp / TRACE-RECURSIVE :ENCAPSULATE NIL
>       Not relevant:        21 tests
>      ok
>      //apparent success (reached end of run-tests.sh normally)
>      Wed Jun  1 11:39:23 EDT 2011
> 
> Thoughts?  Barring objections, I'll try to get this glued together next week.

It seems to be in line with the idea of showing expected failures at
least. It might even provide a little more motivation to interested
parties toward getting the skipped tests enabled :)

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

Re: freeze for 1.0.49

Juho Snellman <jsnell <at> iki.fi> writes:

> 
> I'll be releasing 1.0.49 next weekend. Please let sbcl-devel know of any
regressions.-- Juho Snellman
> 

Has anyone been able to build the current cvs on Win32?  I've tried under cygwin
and mingw, and had the c compiler error out.

gcc -g -Wall -O3 -I. -DSBCL_PREFIX=\"'C:\Program Files/sbcl'\"  -c -o alloc.o al
loc.c
In file included from alloc.c:22:
os.h:137: error: expected '=', ',', ';', 'asm' or '__attribute__' before '*' tok
en
In file included from thread.h:20,
                 from gc-internal.h:20,
                 from alloc.h:17,
                 from alloc.c:23:
...
interrupt.h:110: error: expected specifier-qualifier-list before 'sigset_t'
alloc.c: In function 'pa_alloc':
alloc.c:46: warning: implicit declaration of function 'check_gc_signals_unblocke
d_or_lose'
gmake: *** [alloc.o] Error 1
gmake: Leaving directory `c:/cygwin/usr/local/src/sbcl-dev/sbcl-build/src/runtim
e'

In a related note is anyone building
https://github.com/akovalenko/sbcl-win32-threads tls-63 under Win32 on a regular

Re: freeze for 1.0.49

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

_______________________________________________
Sbcl-devel mailing list
Sbcl-devel <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-devel

Re: freeze for 1.0.49

Elliott Slaughter <elliottslaughter <at> gmail.com> writes:

>     In a related note is anyone building
>     https://github.com/akovalenko/sbcl-win32-threads tls-63 under
>     Win32 on a regular basis?

Well, I am :)

As usually, current URLs for installers and standalone executables are
available at https://github.com/akovalenko/sbcl-win32-threads/wiki .

Additionally, all published builds since 1.0.46.32.263.wth are now
available at http://archive.siftsoft.com/sbcl/ (this link is now
included in the autogenerated wiki page mentioned above).

I publish binaries at the same points when I'm ready to deploy them to
my own customer, so they're generally more well-tested and stable than a
random snapshot from my tree.

P.S. Going to add separate maintainance branches soon (official SBCL
release + MS Windows patchset + occasional regression fixes only).

--

-- 
Regards, Anton Kovalenko
+7(916)345-34-02 | Elektrostal' MO, Russia

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

Re: freeze for 1.0.49

On 2 June 2011 15:15, Anton Kovalenko <anton <at> sw4me.com> wrote:

> I publish binaries at the same points when I'm ready to deploy them to
> my own customer, so they're generally more well-tested and stable than a
> random snapshot from my tree.

Apropos, I think it would be a good idea to promote the windows fork
on sbcl.org, with explanations as its status.

What would be a good link to provde? The forknews one?

What should we say? I think it would be swell if eg. Anton could
provide a short description from his POV, and eg. David -- who's
working on the merge -- could say something about that.

Cheers,

 -- nikodemus

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

Re: [PATCH] Find proper definition source for struct copiers

On 31 May 2011 14:51, Didier Verna <didier <at> lrde.epita.fr> wrote:

> currently sb-introspect correctly sets the definition source for a
> structure constructor, predicate and slot accessors, but not for the
> copier function. The following patch fixes this.

Thanks. I'll merge this as 1.0.49.something.

For future reference (or even now, if you feel up to it): patches as
attachements, as generated by

  git format-patch -1

(or whatever the appropriate number is for the patchset in question)
are strongly preferred. Similarly, we don't use a ChangeLog, but
verbose commit messages -- which git format-patch brings along for the
ride nicely.

Finally, putting the patch up on launchpad with a "review" tag makes
it easier to keep track off -- but sending to sbcl-devel is perfectly
kosher as well. It just makes the patch easier to refer to later.

Cheers,

 -- nikodemus

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger. 
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Data protection magic?
Nope - It's vRanger. Get your free trial download today. 
http://p.sf.net/sfu/quest-sfdev2dev

Re: freeze for 1.0.49

Jim Wise <jwise <at> draga.com> writes:

> I'd like to change this to something like:
>
>      Finished running tests.
>      Status:
>       Expected failure:    packages.impure.lisp / USE-PACKAGE-CONFLICT-SET
>       Expected failure:    packages.impure.lisp / IMPORT-SINGLE-CONFLICT
>       Skipped (broken):    debug.impure.lisp / TRACE-RECURSIVE :ENCAPSULATE NIL
>       Not relevant:        21 tests
>      ok
>      //apparent success (reached end of run-tests.sh normally)
>      Wed Jun  1 11:39:23 EDT 2011
>
> Thoughts?  Barring objections, I'll try to get this glued together next week.

Got a chance to hack on this on the train last night.  Currently looks
like this:

     Finished running tests.
     Status:
      Skipped (broken):    debug.impure.lisp / (TRACE ENCAPSULATE NIL)
      Skipped (broken):    debug.impure.lisp / (TRACE-RECURSIVE ENCAPSULATE NIL)
      Expected failure:    packages.impure.lisp / USE-PACKAGE-CONFLICT-SET
      Expected failure:    packages.impure.lisp / IMPORT-SINGLE-CONFLICT
      (2 tests skipped for this combination of platform and features)
     ok

(This is with just a sample set of tests converted to skip based on
feature checks.)

I'll work on converting the remaining tests which are currently bypassed, and get
this all in after 1.0.49 is cut.  I'll do my best to make the new setup
equivalent to the existing versions, but it would be good if people who
have platforms I don't can make sure the tests still work when I'm done
(I can test SunOS, Darwin, and Linux).

Another issue which could pop up, but shouldn't is that the existing
method (using read-time conditionals to skip tests which would break)
prevents the offending test from being macro-expanded and compiled, as
well as run.  The new code expands and compiles the tests, but skips
them at run-time, updating the tests results.

If there _are_ any tests which will cause problems at expand/compile
time on platforms for which they are currently disabled, this will need
to be fixed.  If you find any, let me know, and I'll be happy to take a look.

Thanks,
--

-- 
				Jim Wise
				jwise <at> draga.com

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Discover what all the cheering's about.
Get your free trial download today. 
http://p.sf.net/sfu/quest-dev2dev2 

_______________________________________________
Sbcl-devel mailing list
Sbcl-devel <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-devel

Re: freeze for 1.0.49

On 2 June 2011 18:53, Jim Wise <jwise <at> draga.com> wrote:

>>       Skipped (broken):    debug.impure.lisp / TRACE-RECURSIVE :ENCAPSULATE NIL
>>       Not relevant:        21 tests

>> Thoughts?  Barring objections, I'll try to get this glued together next week.

I like this!

Cheers,

 -- nikodemus

------------------------------------------------------------------------------
Simplify data backup and recovery for your virtual environment with vRanger.
Installation's a snap, and flexible recovery options mean your data is safe,
secure and there when you need it. Discover what all the cheering's about.
Get your free trial download today. 
http://p.sf.net/sfu/quest-dev2dev2 
_______________________________________________
Sbcl-devel mailing list
Sbcl-devel <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-devel