Bug #132527 for perl5: Bleadperl v5.27.5-398-g19a8de4862 breaks MLEHMANN/AnyEvent-HTTP-2.23.tar.gz

# Fri, 01 Dec 2017 15:27:39 -0800 (Andreas J. Koenig) <andreas.koenig.7os6VVqR@franz.ak.mind.de> - Ticket created [Reply] [Comment]

To:	perlbug [...] perl.org
Date:	Sat, 02 Dec 2017 00:27:18 +0100
From:	Andreas Koenig <andreas.koenig.7os6VVqR [...] franz.ak.mind.de>
Subject:	Bleadperl v5.27.5-398-g19a8de4862 breaks MLEHMANN/AnyEvent-HTTP-2.23.tar.gz

Download (untitled) / with headers
text/plain 4.5k

Thanks to Slaven for discovering this one. bisect points to: : commit 19a8de486236b6004a51aa6c16eb9fcde47d86d1 : Author: Zefram <zefram@fysh.org> : Date: Thu Nov 16 14:56:11 2017 +0000 : : fix lvalue context for 4-arg substr This is the second BBC for this distro in a short row. The other is covered in https://rt.perl.org/Public/Bug/Display.html?id=132142 To produce the new one I applied Ilmari's patch for AnyEvent-7.14. Diagnostics now: : Can't modify delete in substr at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1104, near """)" : Type of arg 1 to AnyEvent::Util::guard must be block or sub {} (not reference constructor) at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1238, near "}" : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1433, near "$ENV{http_proxy" : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1434, near "}" : Compilation failed in require at t/00_load.t line 3. : BEGIN failed--compilation aborted at t/00_load.t line 3. perl -V ------- Summary of my perl5 (revision 5 version 27 subversion 6) configuration: Commit id: 19a8de486236b6004a51aa6c16eb9fcde47d86d1 Platform: osname=linux osvers=4.12.0-2-amd64 archname=x86_64-linux-thread-multi uname='linux k93msid 4.12.0-2-amd64 #1 smp debian 4.12.13-1 (2017-09-19) x86_64 gnulinux ' config_args='-Dprefix=/home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.27.5-398-g19a8de4862/1e0c -Dmyhostname=k93msid -Dinstallusrbinperl=n -Uversiononly -Dusedevel -des -Ui_db -Dlibswanted=cl pthread socket inet nsl gdbm dbm malloc dl ld sun m crypt sec util c cposix posix ucb BSD gdbm_compat -Duseithreads -Uuselongdouble -DEBUGGING=-g' hint=recommended useposix=true d_sigaction=define useithreads=define usemultiplicity=define use64bitint=define use64bitall=define uselongdouble=undef usemymalloc=n default_inc_excludes_dot=define bincompat5005=undef Compiler: cc='cc' ccflags ='-D_REENTRANT -D_GNU_SOURCE -fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2' optimize='-O2 -g' cppflags='-D_REENTRANT -D_GNU_SOURCE -fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include' ccversion='' gccversion='7.2.0' gccosandvers='' intsize=4 longsize=8 ptrsize=8 doublesize=8 byteorder=12345678 doublekind=3 d_longlong=define longlongsize=8 d_longdbl=define longdblsize=16 longdblkind=3 ivtype='long' ivsize=8 nvtype='double' nvsize=8 Off_t='off_t' lseeksize=8 alignbytes=8 prototype=define Linker and Libraries: ld='cc' ldflags =' -fstack-protector-strong -L/usr/local/lib' libpth=/usr/local/lib /usr/lib/gcc/x86_64-linux-gnu/7/include-fixed /usr/include/x86_64-linux-gnu /usr/lib /lib/x86_64-linux-gnu /lib/../lib /usr/lib/x86_64-linux-gnu /usr/lib/../lib /lib libs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc perllibs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc libc=libc-2.24.so so=so useshrplib=false libperl=libperl.a gnulibc_version='2.24' Dynamic Linking: dlsrc=dl_dlopen.xs dlext=so d_dlsymun=undef ccdlflags='-Wl,-E' cccdlflags='-fPIC' lddlflags='-shared -O2 -g -L/usr/local/lib -fstack-protector-strong' Characteristics of this binary (from libperl): Compile-time options: HAS_TIMES MULTIPLICITY PERLIO_LAYERS PERL_COPY_ON_WRITE PERL_DONT_CREATE_GVSV PERL_IMPLICIT_CONTEXT PERL_MALLOC_WRAP PERL_OP_PARENT PERL_PRESERVE_IVUV PERL_USE_DEVEL USE_64_BIT_ALL USE_64_BIT_INT USE_ITHREADS USE_LARGE_FILES USE_LOCALE USE_LOCALE_COLLATE USE_LOCALE_CTYPE USE_LOCALE_NUMERIC USE_LOCALE_TIME USE_PERLIO USE_PERL_ATOF USE_REENTRANT_API Built under linux Compiled at Nov 16 2017 15:19:59 @INC: /home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.27.5-398-g19a8de4862/1e0c/lib/site_perl/5.27.6/x86_64-linux-thread-multi /home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.27.5-398-g19a8de4862/1e0c/lib/site_perl/5.27.6 /home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.27.5-398-g19a8de4862/1e0c/lib/5.27.6/x86_64-linux-thread-multi /home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.27.5-398-g19a8de4862/1e0c/lib/5.27.6 -- andreas

# Fri, 01 Dec 2017 15:27:39 -0800 The RT System itself - Severity low added

# Fri, 01 Dec 2017 15:27:39 -0800 The RT System itself - Type unknown added

# Fri, 01 Dec 2017 18:21:29 -0800 Father Chrysostomos <sprout@cpan.org> - Correspondence added [Reply] [Comment]

RT-Send-CC:

perl5-porters [...] perl.org

Download (untitled) / with headers
text/plain 1.4k

On Fri, 01 Dec 2017 15:27:39 -0800, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote: Show quoted text

> Thanks to Slaven for discovering this one. > > bisect points to: > > : commit 19a8de486236b6004a51aa6c16eb9fcde47d86d1 > : Author: Zefram <zefram@fysh.org> > : Date: Thu Nov 16 14:56:11 2017 +0000 > : > : fix lvalue context for 4-arg substr > > This is the second BBC for this distro in a short row. The other is > covered in https://rt.perl.org/Public/Bug/Display.html?id=132142 > > To produce the new one I applied Ilmari's patch for AnyEvent-7.14. > > Diagnostics now: > > : Can't modify delete in substr at /tmp/loop_over_bdir-15721- > ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1104, near > """)"

I have a fix for that first error. I will push it as soon as tests finish. Show quoted text

> : Type of arg 1 to AnyEvent::Util::guard must be block or sub {} (not > reference constructor) at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent- > HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1238, near "}" > : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23- > 0/blib/lib/AnyEvent/HTTP.pm line 1433, near "$ENV{http_proxy" > : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP-2.23- > 0/blib/lib/AnyEvent/HTTP.pm line 1434, near "}" > : Compilation failed in require at t/00_load.t line 3. > : BEGIN failed--compilation aborted at t/00_load.t line 3.

I hope these are just side-effects of the first error. -- Father Chrysostomos

# Fri, 01 Dec 2017 18:21:30 -0800 The RT System itself - Status changed from 'new' to 'open'

# Fri, 01 Dec 2017 18:48:45 -0800 Father Chrysostomos <sprout@cpan.org> - Correspondence added [Reply] [Comment]

RT-Send-CC:

perl5-porters [...] perl.org

Download (untitled) / with headers
text/plain 1.1k

On Fri, 01 Dec 2017 18:21:29 -0800, sprout wrote: Show quoted text

> On Fri, 01 Dec 2017 15:27:39 -0800, > andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:

> > : Can't modify delete in substr at /tmp/loop_over_bdir-15721- > > ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1104, near > > """)"

> > I have a fix for that first error. I will push it as soon as tests > finish.

Now pushed as 833d07d. Show quoted text

>

> > : Type of arg 1 to AnyEvent::Util::guard must be block or sub {} (not > > reference constructor) at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent- > > HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1238, near "}" > > : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP- > > 2.23- > > 0/blib/lib/AnyEvent/HTTP.pm line 1433, near "$ENV{http_proxy" > > : syntax error at /tmp/loop_over_bdir-15721-ERyp63/AnyEvent-HTTP- > > 2.23- > > 0/blib/lib/AnyEvent/HTTP.pm line 1434, near "}" > > : Compilation failed in require at t/00_load.t line 3. > > : BEGIN failed--compilation aborted at t/00_load.t line 3.

> > I hope these are just side-effects of the first error.

Do you still get these other errors? -- Father Chrysostomos

# Fri, 01 Dec 2017 23:16:51 -0800 (Andreas J. Koenig) <andreas.koenig.7os6VVqR@franz.ak.mind.de> - Correspondence added [Reply] [Comment]

To:	"Father Chrysostomos via RT" <perlbug-followup [...] perl.org>
Date:	Sat, 02 Dec 2017 08:15:16 +0100
Subject:	Re: [perl #132527] Bleadperl v5.27.5-398-g19a8de4862 breaks MLEHMANN/AnyEvent-HTTP-2.23.tar.gz
From:	Andreas Koenig <andreas.koenig.7os6VVqR [...] franz.ak.mind.de>

Download (untitled) / with headers
text/plain 478b

Show quoted text

>>>>> On Fri, 01 Dec 2017 18:48:45 -0800, "Father Chrysostomos via RT" <perlbug-followup@perl.org> said:

Show quoted text

>> I hope these are just side-effects of the first error.

Show quoted text

> Do you still get these other errors?

Beautiful: : All tests successful. : Files=3, Tests=28, 1 wallclock secs ( 0.03 usr 0.02 sys + 0.16 cusr 0.01 csys = 0.22 CPU) : Result: PASS : MLEHMANN/AnyEvent-HTTP-2.23.tar.gz : /usr/bin/make test -- OK This ticket can be closed. Thanks! -- andreas

# Sat, 02 Dec 2017 02:45:19 -0800 Dave Mitchell <davem@iabyn.com> - Status changed from 'open' to 'resolved'

# Mon, 04 Dec 2017 09:37:56 -0800 Zefram <zefram@fysh.org> - Correspondence added [Reply] [Comment]

Date:	Mon, 4 Dec 2017 17:37:42 +0000
To:	perl5-porters [...] perl.org
Subject:	Re: [perl #132527] Bleadperl v5.27.5-398-g19a8de4862 breaks MLEHMANN/AnyEvent-HTTP-2.23.tar.gz
From:	Zefram <zefram [...] fysh.org>

Download (untitled) / with headers
text/plain 1k

Father Chrysostomos via RT wrote: Show quoted text

>On Fri, 01 Dec 2017 15:27:39 -0800, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:

>> : Can't modify delete in substr at /tmp/loop_over_bdir-15721- >> ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1104, near >> """)"

> >I have a fix for that first error. I will push it as soon as tests finish.

The change you've made is to use loose lvalue context for the first arg of a 4-arg substr. I think that is a bad change and should be reverted. 4-arg substr does not only sometimes mutate its first argument; it *always* writes to it. Doing so is the primary purpose of 4-arg substr. It should therefore require a fully-qualified lvalue. It is correct for it to reject the use of a delete there, consistent with how assignment to a delete is rejected. As far as I can see, AnyEvent::HTTP has no reason to use a mutating substr on a delete there. It should be using either a mutating substr on a hash element left in place, or a non-mutating substr on a delete. I reckon commit 833d07d41793fe678f93433fca9520ec1a22f750 should be reverted. -zefram

# Mon, 04 Dec 2017 11:41:46 -0800 Father Chrysostomos <sprout@cpan.org> - Correspondence added [Reply] [Comment]

RT-Send-CC:

perl5-porters [...] perl.org

Download (untitled) / with headers
text/plain 2.6k

On Mon, 04 Dec 2017 09:37:56 -0800, zefram@fysh.org wrote: Show quoted text

> Father Chrysostomos via RT wrote:

> > On Fri, 01 Dec 2017 15:27:39 -0800, > > andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:

> >> : Can't modify delete in substr at /tmp/loop_over_bdir-15721- > >> ERyp63/AnyEvent-HTTP-2.23-0/blib/lib/AnyEvent/HTTP.pm line 1104, > >> near > >> """)"

> > > > I have a fix for that first error. I will push it as soon as tests > > finish.

> > The change you've made is to use loose lvalue context for the first > arg > of a 4-arg substr. I think that is a bad change and should be > reverted. > 4-arg substr does not only sometimes mutate its first argument; it > *always* writes to it. Doing so is the primary purpose of 4-arg > substr. > It should therefore require a fully-qualified lvalue. It is correct > for > it to reject the use of a delete there, consistent with how assignment > to a delete is rejected.

The acceptance or rejection of any particular op as an lvalue is necessarily arbitrary. There is actually no syntactic need for 3=4 to be prohibited, because it will croak at run time anyway. The only reason it is prohibited is that it would be unhelpful to programmers to allow such obviously sloppy code to compile. This arbitrariness actually prevents some useful constructs: ++tie $X, "foo"; If foo::TIESCALAR returns an object with overloading that makes ++ meaningful, then that construct could be used, were it not for the compile-time check. I’m not arguing that this case should be made to work. I’m just using it as an example to demonstrate how arbitrary compile-time lvalue checks are. Other examples: ${\3} = 4; # This could die at compile time. $x.$y = 7; # SomeClass::(. could be an lvalue function. for(3){$_++} # This could die at compile time. Show quoted text

> As far as I can see, AnyEvent::HTTP has no reason to use a mutating > substr on a delete there. It should be using either a mutating substr > on a hash element left in place, or a non-mutating substr on a delete.

It may be that AnyEvent::HTTP is using four-argument substr unnecessarily. But I tend to see *any* CPAN failure as merely a canary in the coalmine, and not the only failure. The fact that AnyEvent::HTTP uses it that way is enough to prove that we have broken backward-compatibility when we did not need to. Since the first argument to substr has never ever had strict compile-time lvalue checks applied to it, I think it would be wrong to impose it now, because of the arbitrariness I mentioned above. Show quoted text

> I reckon commit 833d07d41793fe678f93433fca9520ec1a22f750 should be > reverted.

I think backward compatibility should win. The current (restored) behaviour is not clearly a bug. -- Father Chrysostomos

# Mon, 04 Dec 2017 13:53:36 -0800 Zefram <zefram@fysh.org> - Correspondence added [Reply] [Comment]

From:	Zefram <zefram [...] fysh.org>
Subject:	Re: [perl #132527] Bleadperl v5.27.5-398-g19a8de4862 breaks MLEHMANN/AnyEvent-HTTP-2.23.tar.gz
To:	perl5-porters [...] perl.org
Date:	Mon, 4 Dec 2017 21:53:25 +0000

Download (untitled) / with headers
text/plain 579b

Father Chrysostomos via RT wrote: Show quoted text

>The acceptance or rejection of any particular op as an lvalue is >necessarily arbitrary.

This is true; there is some arbitrariness, and even a few cases where it's useful to work around a forbidden type of lvalue. But the arbitrariness is mostly consistent. For it to be inconsistent here seems to be a bug. Show quoted text

>The only reason it is prohibited is that it would be unhelpful to >programmers to allow such obviously sloppy code to compile.

It is just as helpful to reject substr on a delete as it is to reject assignment to a delete. -zefram

# Mon, 04 Dec 2017 15:26:33 -0800 Father Chrysostomos <sprout@cpan.org> - Correspondence added [Reply] [Comment]

RT-Send-CC:

perl5-porters [...] perl.org

Download (untitled) / with headers
text/plain 894b

On Mon, 04 Dec 2017 13:53:36 -0800, zefram@fysh.org wrote: Show quoted text

> Father Chrysostomos via RT wrote:

> >The acceptance or rejection of any particular op as an lvalue is > >necessarily arbitrary.

> > This is true; there is some arbitrariness, and even a few cases where it's > useful to work around a forbidden type of lvalue. But the arbitrariness > is mostly consistent. For it to be inconsistent here seems to be a bug. >

> >The only reason it is prohibited is that it would be unhelpful to > >programmers to allow such obviously sloppy code to compile.

> > It is just as helpful to reject substr on a delete as it is to reject > assignment to a delete.

And it is also helpful to allow it. Traditionally Perl has retained backward compatibility in such cases. I do not believe we can reach an agreement on this. It would be helpful if someone else would chip in. -- Father Chrysostomos