git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, Eric Sunshine <sunshine@sunshineco.com>,
	Sam James <sam@gentoo.org>, Andreas Larsson <andreas@gaisler.com>
Subject: Re: [PATCH] chainlint.pl: Extend regexp pattern for /proc/cpuinfo on Linux SPARC
Date: Mon, 20 May 2024 18:48:44 +0200	[thread overview]
Message-ID: <503a99f3511559722a3eeef15d31027dfe617fa1.camel@physik.fu-berlin.de> (raw)
In-Reply-To: <xmqqseyclaq8.fsf@gitster.g>

Hi Junio,

On Mon, 2024-05-20 at 09:16 -0700, Junio C Hamano wrote:
> John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> writes:
> 
> > On SPARC systems running Linux, individual processors are denoted with
> > "CPUnn:" in /proc/cpuinfo instead of the usual "processor NN:" so that
> > the current regexp in ncores() returns 0. Extend the regexp to match
> > lines with "CPUnn:" as well to properly detect the number of available
> > cores on these systems.
> > 
> > Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
> > ---
> >  t/chainlint.pl | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/t/chainlint.pl b/t/chainlint.pl
> > index 556ee91a15..63cac942ac 100755
> > --- a/t/chainlint.pl
> > +++ b/t/chainlint.pl
> > @@ -718,7 +718,7 @@ sub ncores {
> >  	# Windows
> >  	return $ENV{NUMBER_OF_PROCESSORS} if exists($ENV{NUMBER_OF_PROCESSORS});
> >  	# Linux / MSYS2 / Cygwin / WSL
> > -	do { local @ARGV='/proc/cpuinfo'; return scalar(grep(/^processor[\s\d]*:/, <>)); } if -r '/proc/cpuinfo';
> > +	do { local @ARGV='/proc/cpuinfo'; return scalar(grep(/^processor[\s\d]*:||^CPU[\d]*:/, <>)); } if -r '/proc/cpuinfo';
> 
> Is the doubled || intended?  Doesn't it introduce an empty pattern
> that slurps every single line of /proc/cpuinfo?

I'm not a Perl expert by any means, so I wasn't sure what the correct logical OR
operator would be. If it turns out to be wrong, let's fix that.

> I was wondering if we want to first add the "reasonable fallback"
> Eric mentioned ealier, and then build on top, whose result may look
> like the attached.  You can enable the STDERR thing with your double
> "||" added back and see what "cd t && perl chainlint.pl" produces.
> 
> Thanks.
> 
> diff --git i/t/chainlint.pl w/t/chainlint.pl
> index 556ee91a15..775f06281b 100755
> --- i/t/chainlint.pl
> +++ w/t/chainlint.pl
> @@ -718,7 +718,13 @@ sub ncores {
>  	# Windows
>  	return $ENV{NUMBER_OF_PROCESSORS} if exists($ENV{NUMBER_OF_PROCESSORS});
>  	# Linux / MSYS2 / Cygwin / WSL
> -	do { local @ARGV='/proc/cpuinfo'; return scalar(grep(/^processor[\s\d]*:/, <>)); } if -r '/proc/cpuinfo';
> +	do {
> +		local @ARGV='/proc/cpuinfo';
> +		my @num = grep(/^processor[\s\d]*:|^CPU[\d]*:/, <>);
> +# print STDERR "FOUND <@num>\n";
> +		return 1 if (!@num);
> +		return scalar(@num);
> +	} if -r '/proc/cpuinfo';
>  	# macOS & BSD
>  	return qx/sysctl -n hw.ncpu/ if $^O =~ /(?:^darwin$|bsd)/;
>  	return 1;

This seems to work fine for me as well. If you post it as a patch, I'm more than
happy to give it a Tested-By.

Btw, it would be great if this could be extended to support the output for the
Alpha architecture as well since the testsuite fails the same way [1]. The output
for /proc/cpuinfo looks like this [2]:

(alpha-chroot)root@p100:/# cat /proc/cpuinfo
cpu                     : Alpha
cpu model               : ev67
cpu variation           : 0
cpu revision            : 0
cpu serial number       : JA00000000
system type             : QEMU
system variation        : QEMU_v8.0.92
system revision         : 0
system serial number    : AY00000000
cycle frequency [Hz]    : 250000000
timer frequency [Hz]    : 250.00
page size [bytes]       : 8192
phys. address bits      : 44
max. addr. space #      : 255
BogoMIPS                : 2500.00
platform string         : AlphaServer QEMU user-mode VM
cpus detected           : 8
cpus active             : 4
cpu active mask         : 0000000000000095
L1 Icache               : n/a
L1 Dcache               : n/a
L2 cache                : n/a
L3 cache                : n/a

Thanks so much for helping with the fix!

Adrian

> [1] https://buildd.debian.org/status/fetch.php?pkg=git&arch=alpha&ver=1%3A2.45.1-1&stamp=1716194983&raw=0
> [2] https://lore.kernel.org/all/20230901204251.137307-4-richard.henderson@linaro.org/

-- 
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer
`. `'   Physicist
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913

  reply	other threads:[~2024-05-20 16:48 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-20 11:11 [PATCH] chainlint.pl: Extend regexp pattern for /proc/cpuinfo on Linux SPARC John Paul Adrian Glaubitz
2024-05-20 16:16 ` Junio C Hamano
2024-05-20 16:48   ` John Paul Adrian Glaubitz [this message]
2024-05-20 16:52     ` Eric Sunshine
2024-05-20 16:50   ` Eric Sunshine
2024-05-20 17:07     ` Eric Sunshine
2024-05-20 17:23     ` Junio C Hamano
2024-05-20 19:01 ` [PATCH 0/3] improve chainlint.pl CPU count computation Eric Sunshine
2024-05-20 19:01   ` [PATCH 1/3] chainlint.pl: make CPU count computation more robust Eric Sunshine
2024-05-20 19:01   ` [PATCH 2/3] chainlint.pl: fix incorrect CPU count on Linux SPARC Eric Sunshine
2024-05-22  8:32     ` Carlo Marcelo Arenas Belón
2024-05-22  8:47       ` John Paul Adrian Glaubitz
2024-05-22  9:05         ` Eric Sunshine
2024-05-22 19:00           ` Junio C Hamano
2024-05-22 19:11             ` Eric Sunshine
2024-05-27 19:48               ` John Paul Adrian Glaubitz
2024-05-27 20:12                 ` Eric Sunshine
2024-05-20 19:01   ` [PATCH 3/3] chainlint.pl: latch CPU count directly reported by /proc/cpuinfo Eric Sunshine
2024-05-20 19:17   ` [PATCH 0/3] improve chainlint.pl CPU count computation John Paul Adrian Glaubitz
2024-05-20 19:19     ` Eric Sunshine
2024-05-20 19:23       ` John Paul Adrian Glaubitz
2024-05-21 14:28         ` John Paul Adrian Glaubitz
2024-05-21 16:18           ` Eric Sunshine

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=503a99f3511559722a3eeef15d31027dfe617fa1.camel@physik.fu-berlin.de \
    --to=glaubitz@physik.fu-berlin.de \
    --cc=andreas@gaisler.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=sam@gentoo.org \
    --cc=sunshine@sunshineco.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).