Darwin 8.6.0 (Mac OS X 10.4.6)
Dominic Dunlop
Are there any objections to or suggestions concerning me starting to
send in smokes of 32-bit and 64-bit builds of bleadperl on alternate
nights? (My system's not beefy enough to get through the lot every
night.) The reports are currently as below. One thing that concerns
me is that the Subject is the same in both cases. Should I tweak it?
How much?
------------------------------------------------------------------------
-------------
From: should...@mac.com
Subject: Smoke [5.9.4] 28148 PASS darwin 8.6.0 (macppc/1 cpu)
Date: 10 May 2006 17:19:00.000
To: domo
Automated smoke report for 5.9.4 patch 28148
Tullamore.local: PowerMac8,1 (1600 MHz) (macppc/1 cpu)
on darwin - 8.6.0
using cc version 4.0.1 (Apple Computer, Inc. build 5250)
smoketime 4 hours 58 minutes (average 1 hour 14 minutes)
Summary: PASS
O = OK F = Failure(s), extended report at the bottom
X = Failure(s) under TEST but not under harness
? = still running or test results not (yet) available
Build failures during: - = unknown or N/A
c = Configure, m = make, M = make (after miniperl), t = make test-prep
28148 Configuration (common) -D'optimize=-O1 -g'
----------- ---------------------------------------------------------
O O O O -Duse64bitint
O O O O -Duse64bitint -Duseithreads
| | | +----- PERLIO = perlio -DDEBUGGING
| | +------- PERLIO = stdio -DDEBUGGING
| +--------- PERLIO = perlio
+----------- PERLIO = stdio
Locally applied patches:
SMOKE28148
Compiler messages(gcc):
regcomp.c: In function 'S_study_chunk':
regcomp.c:1563: warning: comparison is always false due to limited
range of data type
sv.c: In function 'S_more_bodies':
sv.c:1067: warning: format '%d' expects type 'int', but argument 5
has type 'size_t'
sv.c:1067: warning: format '%d' expects type 'int', but argument 7
has type 'size_t'
sv.c:1067: warning: format '%d' expects type 'int', but argument 8
has type 'long unsigned int'
sv.c: In function 'Perl_sv_upgrade':
sv.c:1255: warning: comparison is always false due to limited range
of data type
sv.c:1309: warning: comparison is always false due to limited range
of data type
pp_sys.c:309: warning: 'S_emulate_eaccess' defined but not used
byterun.c: In function 'byterun':
byterun.c:894: warning: comparison is always false due to limited
range of data type
re_comp.c: In function 'S_study_chunk':
re_comp.c:1563: warning: comparison is always false due to limited
range of data type
sv.c: In function 'Perl_sv_dup':
sv.c:10026: warning: comparison is always false due to limited range
of data type
--
Report by Test::Smoke v1.19_76 build 914 running on perl 5.8.6
(Reporter v0.026 / Smoker v0.027)
------------------------------------------------------------------------
-------------
From: should...@mac.com
Subject: Smoke [5.9.4] 28126 PASS darwin 8.6.0 (macppc/1 cpu)
Date: 9 May 2006 02:31:00.000
To: do...@pt.lu
Automated smoke report for 5.9.4 patch 28126
Tullamore.local: PowerMac8,1 (1600 MHz) (macppc/1 cpu)
on darwin - 8.6.0
using cc version 4.0.1 (Apple Computer, Inc. build 5250)
smoketime 5 hours 24 minutes (average 1 hour 21 minutes)
Summary: PASS
O = OK F = Failure(s), extended report at the bottom
X = Failure(s) under TEST but not under harness
? = still running or test results not (yet) available
Build failures during: - = unknown or N/A
c = Configure, m = make, M = make (after miniperl), t = make test-prep
28126 Configuration (common) none <== I'll be adding "-
D'optimize=-O1 -g'"" -O3 takes too long
----------- ---------------------------------------------------------
O O O O
O O O O -Duseithreads
| | | +----- PERLIO = perlio -DDEBUGGING
| | +------- PERLIO = stdio -DDEBUGGING
| +--------- PERLIO = perlio
+----------- PERLIO = stdio
Locally applied patches:
SMOKE28126
Compiler messages(gcc):
regcomp.c: In function 'S_study_chunk':
regcomp.c:1563: warning: comparison is always false due to limited
range of data type
sv.c: In function 'Perl_sv_upgrade':
sv.c:1255: warning: comparison is always false due to limited range
of data type
sv.c:1309: warning: comparison is always false due to limited range
of data type
pp_sys.c:309: warning: 'S_emulate_eaccess' defined but not used
byterun.c: In function 'byterun':
byterun.c:894: warning: comparison is always false due to limited
range of data type
re_comp.c: In function 'S_study_chunk':
re_comp.c:1563: warning: comparison is always false due to limited
range of data type
sv.c: In function 'S_more_bodies':
sv.c:1067: warning: format '%d' expects type 'int', but argument 5
has type 'size_t'
sv.c:1067: warning: format '%d' expects type 'int', but argument 7
has type 'size_t'
sv.c:1067: warning: format '%d' expects type 'int', but argument 8
has type 'long unsigned int'
sv.c: In function 'Perl_sv_dup':
sv.c:10026: warning: comparison is always false due to limited range
of data type
--
Report by Test::Smoke v1.19_76 build 914 running on perl 5.8.6
(Reporter v0.026 / Smoker v0.027)
--
Dominic Dunlop
H.Merijn Brand
wrote:
> Hoping that this list is not utterly dead...
No it isn't :)
But it is, err, very low volume
> Are there any objections to or suggestions concerning me starting to
> send in smokes of 32-bit and 64-bit builds of bleadperl on alternate
> nights?
No, au contraire. All smokes are welcome!
My HP-UX 10.20 also cannot finish in 24 hours every time.
> (My system's not beefy enough to get through the lot every
> night.) The reports are currently as below. One thing that concerns
> me is that the Subject is the same in both cases. Should I tweak it?
Isn't it possible to get those in one configuration, and just let the smoke
take twice as long?
> How much?
If you can tweak it, have a look in Test/SysInfo.pm
> ------------------------------------------------------------------------
--
H.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)
using & porting perl 5.6.2, 5.8.x, 5.9.x on HP-UX 10.20, 11.00, 11.11,
& 11.23, SuSE 10.0, AIX 4.3 & 5.2, and Cygwin. http://qa.perl.org
http://mirrors.develooper.com/hpux/ http://www.test-smoke.org
http://www.goldmark.org/jeff/stupid-disclaimers/
Dominic Dunlop
> Isn't it possible to get [32- and 64-bit builds] in one
> configuration, and just let the smoke
> take twice as long?
>
I'd rather not: I use the system for real work during the day, and
having smokes running at the same time -- even nice smokes -- slows
things down. Also, while I did test runs of smokes with the Mac's
"Energy saver" settings tweaked for maximum performance, I won't be
doing that for the smokes run from cron, as it makes the system
rather hot and noisy. This means the smokes I'll be submitting will
take longer than the ones I pasted into my mail.
> If you can tweak it, have a look in Test/SysInfo.pm
Right. I'll change the subject line so that it reads
Subject: Smoke [5.9.4] 28148 PASS darwin 8.6.0 (macppc/1 cpu) [32-bit]
or similar, and see if anybody (or anything) squawks.
--
Dominic Dunlop
Abe Timmerman
> On Wed, 10 May 2006 21:06:04 +0200, Dominic Dunlop <should...@mac.com>
>
> wrote:
> > Hoping that this list is not utterly dead...
>
> No it isn't :)
> But it is, err, very low volume
for any value of low close to none :-)
> If you can tweak it, have a look in Test/SysInfo.pm
>
> > ------------------------------------------------------------------------
> > From: should...@mac.com
> > Subject: Smoke [5.9.4] 28148 PASS darwin 8.6.0 (macppc/1 cpu)
> > Date: 10 May 2006 17:19:00.000
> > To: domo
> >
> > Automated smoke report for 5.9.4 patch 28148
> > Tullamore.local: PowerMac8,1 (1600 MHz) (macppc/1 cpu)
If you know of a way to translate that "PowerMacX,Y" to something like G[345]
(without hardcoding it), that would be nice.
And if this is a G5, can you also plan a smoke for -Duse64bitall?
Good luck,
Abe
--
Adding a -w warning for the "pseudohashes are going away, repent and give away
your all wordly possessions" to 5.8.1 is okay, but what I don't understand is
why Sarathy is thinking people would be using -w that much more than they
would be reading the 5.8.0 announcement? :-)
-- Jarkko Hietaniemi on p5p @ 2002-09-06
Dominic Dunlop
> If you know of a way to translate that "PowerMacX,Y" to something
> like G[345]
> (without hardcoding it), that would be nice.
How about
$ system_profiler SPHardwareDataType
Hardware:
Hardware Overview:
Machine Name: iMac G5
Machine Model: PowerMac8,1
CPU Type: PowerPC G5 (3.0)
Number Of CPUs: 1
CPU Speed: 1.6 GHz
L2 Cache (per CPU): 512 KB
Memory: 1 GB
Bus Speed: 533 MHz
Boot ROM Version: 5.2.2f2
Serial Number: W84460SEPP6
I'm looking into patching Test::Smoke::Sysinfo to get better platform-
specific information by parsing this output (or the corresponding XML
-- don't worry: I won't haul in modules to do it). How would you like
fields from the above to be presented on the Subject line?
>
> And if this is a G5, can you also plan a smoke for -Duse64bitall?
Yes, that was the point of the original mail: because I want things
to be out of the way by the time I start using the system in anger
each day, I'll be running 32-bit and 64-bit smokes on alternate
nights. If I can, I'll find a way of distinguishing between the two
in the reports' subject lines.
--
Dominic Dunlop
Dominic Dunlop
On 2006–05–13, at 17:44, Abe Timmerman wrote:
> Op een mooie lentedag (Thursday 11 May 2006 23:09),schreef
> H.Merijn Brand:
>> If you can tweak it, have a look in Test/SysInfo.pm
>>
>>> --------------------------------------------------------------------
>>> ----
>>> From: should...@mac.com
>>> Subject: Smoke [5.9.4] 28148 PASS darwin 8.6.0 (macppc/1 cpu)
>>> Date: 10 May 2006 17:19:00.000
>>> To: domo
>>>
>>> Automated smoke report for 5.9.4 patch 28148
>>> Tullamore.local: PowerMac8,1 (1600 MHz) (macppc/1 cpu)
>
> If you know of a way to translate that "PowerMacX,Y" to something
> like G[345]
> (without hardcoding it), that would be nice.
OK. I've done some tweaking:
Darwin( )
If the system_profiler program is accessible (meaning that
this is Mac
OS X), use it to find information; otherwise treat as "BSD".
Which gives this on G3, G4 and G5 mono-processors
domo@Dave...:smoke$ perl -I. -MTest::Smoke::SysInfoNew -lwe 'print
"$_: ", Test::Smoke::SysInfo->new->$_ foreach qw(cpu_type cpu ncpu
host os)'
cpu_type: ppc750
cpu: iMac (400 MHz)
ncpu: 1
host: Daventry.local
os: darwin - 8.6.0 (Mac OS X)
domo@Jung...:smoke$ perl -I. -MTest::Smoke::SysInfoNew -lwe 'print
"$_: ", Test::Smoke::SysInfo->new->$_ foreach qw(cpu_type cpu ncpu
host os)'
cpu_type: ppcG4
cpu: PowerBook G4 12" (867 MHz)
ncpu: 1
host: Junglinster.local
os: darwin - 8.6.0 (Mac OS X)
domo@Tull...:smoke$ perl -I. -MTest::Smoke::SysInfoNew -lwe 'print
"$_: ", Test::Smoke::SysInfo->new->$_ foreach qw(cpu_type cpu ncpu
host os)'
cpu_type: ppcG5
cpu: iMac G5 1600 MHz
ncpu: 1
host: Tullamore.local
os: darwin - 8.6.0 (Mac OS X)
Will that do?
Not having an Intel-based Mac to hand, I've not attempted to prettify
whatever it is that system_profiler delivers on that platform -- see
the example here: <http://appleintelfaq.com/imac/
system_profiler.html>. In particular, T::S::SysInfo currently has
nothing to say about the number of cores, as opposed to the number of
processors.
Abe Timmerman
> On 2006–05–13, at 17:44, Abe Timmerman wrote:
> > If you know of a way to translate that "PowerMacX,Y" to something
> > like G[345]
> > (without hardcoding it), that would be nice.
>
> How about
>
> $ system_profiler SPHardwareDataType
> Hardware:
>
> Hardware Overview:
>
> Machine Name: iMac G5
> Machine Model: PowerMac8,1
> CPU Type: PowerPC G5 (3.0)
> Number Of CPUs: 1
> CPU Speed: 1.6 GHz
Brill! I'll deal with that later on in the week.
[snip]
> > And if this is a G5, can you also plan a smoke for -Duse64bitall?
>
> Yes, that was the point of the original mail: because I want things
> to be out of the way by the time I start using the system in anger
> each day, I'll be running 32-bit and 64-bit smokes on alternate
> nights. If I can, I'll find a way of distinguishing between the two
> in the reports' subject lines.
IMO it is not important, we don't really distinguish between the different
Pentiums in the subject line. -Duse64bitall is just another way of
configuring perl, that doesn't change your machine/platform.
HTH +
Good luck,
Abe
--
Ahh, duh. (Thumping forehead against keybsalf<9iw.)
-- Jarkko Hietaniemi on p5p @ 2003-02-11
Dominic Dunlop
> Brill! I'll deal with that later on in the week.
OK. I've attached the patch for SysInfo.pm. I have not done anything
about the version, attribution or copyright.
--
Dominic Dunlop
Abe Timmerman
Oh cheers! I'll take care of the rest.
> --- Test/Smoke/SysInfo.pm-as-received 2005-09-11 01:14:13.000000000 +0200
> +++ Test/Smoke/SysInfo.pm 2006-05-15 15:22:33.000000000 +0200
[snip]
> + $system_profiler{'CPU Type'} =~ s/PowerPC\s*(\w+).*/ppc$1/;
I've changed that to macppc. Should we change the '750' into G3?
> + $system_profiler{'CPU Speed'} =~
> + s/(\d+(?:\.\d+)?)\s*GHz/sprintf("%d MHz", $1 * 1000)/e;
> +
> + return {
> + _cpu_type => $system_profiler{'CPU Type'} || __get_cpu_type(),
> + _cpu => "$system_profiler{'Machine Name'}" .
My iBook G4/10.3 only seems to have 'Machine Model', so I've tweaked that a
bit.
> + " ($system_profiler{'CPU Speed'})" || __get_cpu,
> + _ncpu => $system_profiler{'Number Of CPUs'},
> + _host => __get_hostname(),
> + _os => __get_os() . " (Mac OS X)",
Thanks applied as change #965.
Good luck,
Abe
--
"Jarkko Hietaniemi" is actually the code name for a whole team of Finnish
super-programmers, capable of working continuously 25 hours a day without
tripping each other up, and running solely only on intravenous caffeine.
-- Nicholas Clark on p5p @ 2002-03-04
Dominic Dunlop
On 2006–05–16, at 10:29, Abe Timmerman wrote:
> I've changed that to macppc.
Fine.
> Should we change the '750' into G3?
Your call. As far as I can tell (and looking at <http://
www.everymac.com/systems/apple/index.html>) every Mac billed by Apple
as having a G3 processor used a PPC 750 CPU. Some processor upgrades
use a 755, so, to be safe, it might be worth replacing "PowerPC 75
[05]" with ppcG3.
--
Dominic Dunlop