Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On Thu, Aug 20, 2015 at 10:23 PM, Peter Jeremy pe...@rulingia.com wrote: On my E3-1230v2 system (not a slouch), the difference is 1s to 110s - which is in line with Matthew's results. Yes, this change makes ports almost unusable on FreeBSD 9.3-stable. A two-orders-of-magnitude slowdown in make startup is unreasonable and I formally request r394573 be backed out. I can see the benefits of the functionality but the current cost is too high. Thanks, I'll be eagerly awaiting this change. -- Regards, Torfinn Ingolfsen ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
OTOH I do not want to use ghostscript ! In my view ghost is a critically important port. I will not be able to use FreeBSD productively withouth it. Since I have no time or skill to contribute to address the slowness problem, I'm happy to put up with the extra hour. Anton ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
From m...@freebsd.org Thu Aug 20 09:33:38 2015 +--On 20 ao=C3=BBt 2015 08:24:25 +0100 Anton Shterenlikht = me...@bris.ac.uk wrote: | OTOH I do not want to use ghostscript ! |=20 | In my view ghost is a critically important port. | I will not be able to use FreeBSD productively withouth it. | Since I have no time or skill to contribute to | address the slowness problem, I'm happy to put | up with the extra hour. I'm not exactly sure what you're all complaining about. The port just works fine, it does take a few seconds for make to iterate over all options, but it's because the port has way too many options. It's not as if you're all sitting behind your desk looking at the screen waiting for it to finish. I think something mush have changed recently in ghostscript. If I run e.g. portmaster -L, I get quick progress to ghostcript, where portmaster might pause for over a minute. Perhaps the number of options increased? The extra hour is an exaggeration on my part, and I apologise for it. I haven't built ghostscript from ports for a long time now, so cannot comment on the build times. Anton ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On 2015-08-20 09:32, Mathieu Arnold wrote: +--On 20 août 2015 08:24:25 +0100 Anton Shterenlikht me...@bris.ac.uk wrote: | OTOH I do not want to use ghostscript ! | | In my view ghost is a critically important port. | I will not be able to use FreeBSD productively withouth it. | Since I have no time or skill to contribute to | address the slowness problem, I'm happy to put | up with the extra hour. I'm not exactly sure what you're all complaining about. The port just works fine, it does take a few seconds for make to iterate over all options, but it's because the port has way too many options. It's not as if you're all sitting behind your desk looking at the screen waiting for it to finish. Perhaps some small statistics for a port with 124 options (www/apache24). With the given example commands it should be not hard to verify the results Test was done on a 8 core (3GHz) system with fast SSD's and 24GB RAM $ svn log -q -l 10 /usr/ports/Mk r393817 | netchild | 2015-08-09 21:14:13 +0200 (Sun, 09 Aug 2015) = OK r393878 | antoine | 2015-08-10 19:06:50 +0200 (Mon, 10 Aug 2015) = OK r393903 | jbeich | 2015-08-10 22:14:16 +0200 (Mon, 10 Aug 2015) = OK r393984 | jbeich | 2015-08-12 01:20:48 +0200 (Wed, 12 Aug 2015) = OK r394258 | feld | 2015-08-14 22:59:29 +0200 (Fri, 14 Aug 2015) = OK r394503 | mat | 2015-08-17 15:31:25 +0200 (Mon, 17 Aug 2015) = OK (log grow with additional ~100.000 lines, slower) r394508 | mat | 2015-08-17 16:20:40 +0200 (Mon, 17 Aug 2015) = OK r394569 | mat | 2015-08-18 12:39:07 +0200 (Tue, 18 Aug 2015) = OK r394572 | rakuco | 2015-08-18 12:51:01 +0200 (Tue, 18 Aug 2015) = OK r394573 | mat | 2015-08-18 13:00:57 +0200 (Tue, 18 Aug 2015) = BROKEN explode r394770 | mat | 2015-08-19 11:28:06 +0200 (Wed, 19 Aug 2015) = BROKEN explode $ cd www/apache24 $ svn up -$testrev /usr/ports/Mk $ script -q DEBUG.make.$rev make -dA -V PKGNAME *1.3G Aug 20 05:52 DEBUG.make.r394770 = manual break send after 2 min.* *32M Aug 20 06:24 DEBUG.make.r394573 = manual break send after 15 sec.* 16M Aug 20 06:15 DEBUG.make.r393878 = finished after 4 sec. 16M Aug 20 06:17 DEBUG.make.r394258 = finished after 4 sec. 29M Aug 20 06:19 DEBUG.make.r394503 = finished after 4 sec. 29M Aug 20 06:21 DEBUG.make.r394508 = finished after 4 sec. 29M Aug 20 06:22 DEBUG.make.r394569 = finished after 4 sec. 29M Aug 20 06:23 DEBUG.make.r394572 = finished after 4 sec. $ wc -l DEBUG.make* 569729 DEBUG.make.r394770 = not finished after 2 min, size 1.x GB 242899 DEBUG.make.r393878 242899 DEBUG.make.r394258 348109 DEBUG.make.r394503 348141 DEBUG.make.r394508 350125 DEBUG.make.r394569 350125 DEBUG.make.r394572 38521 DEBUG.make.r394573 = stopped after 15 sec. 2490548 total ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On Thu, 20 Aug 2015, Mathieu Arnold wrote: I'm not exactly sure what you're all complaining about. The port just works fine, it does take a few seconds for make to iterate over all options, but it's because the port has way too many options. It's not as if you're all sitting behind your desk looking at the screen waiting for it to finish. Ok , sitting it out... [portsnap fetch update, portmaster -va] === All ports are up to date [root@mbox ~]# cd /usr/ports/www/apache24 [root@mbox /usr/ports/www/apache24]# time date Thu Aug 20 15:16:17 CEST 2015 real0m0.004s user0m0.000s sys 0m0.008s [root@mbox /usr/ports/www/apache24]# time make clean === Cleaning for apache24-2.4.16_1 real2m40.647s user2m39.589s sys 0m0.688s [root@mbox /usr/ports/www/apache24]# time date Thu Aug 20 15:19:08 CEST 2015 real0m0.003s user0m0.002s sys 0m0.000s Yeah \0/ finished ! But ''find /usr/ports -name work -ls'' is faster... (~1000 ports) Regards, Henk ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On Thu, 20 Aug 2015, Mathieu Arnold wrote: [...] It's not as if you're all sitting behind your desk looking at the screen waiting for it to finish. I can not wait for this: ls -al /usr/ports/print/ghostscript9 total 88 drwxr-xr-x3 root wheel512 Aug 20 12:15 . drwxr-xr-x 262 root wheel 5632 Aug 19 15:57 .. -rw-r--r--1 root wheel 5358 Jun 22 20:33 Makefile -rw-r--r--1 root wheel 23028 Jan 22 2014 Makefile.drivers -rw-r--r--1 root wheel317 Jan 22 2014 distinfo drwxr-xr-x2 root wheel 1536 Jun 22 22:44 files -rw-r--r--1 root wheel 1038 Jan 22 2014 pkg-descr -rw-r--r--1 root wheel 29669 Nov 13 2014 pkg-plist cd /usr/ports/print/ghostscript9 [root@mbox /usr/ports/print/ghostscript9]# time date Thu Aug 20 13:49:43 CEST 2015 real0m0.003s user0m0.003s sys 0m0.001s [root@mbox /usr/ports/print/ghostscript9]# time make clean ^C real71m50.653s user71m5.993s sys 0m0.888s [root@mbox /usr/ports/print/ghostscript9]# time date Thu Aug 20 15:01:46 CEST 2015 real0m0.004s user0m0.003s sys 0m0.000s ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On 2015-Aug-20 10:08:37 -0500, Matthew D. Fuller fulle...@over-yonder.net wrote: On Thu, Aug 20, 2015 at 08:32:07AM +0100 I heard the voice of Mathieu Arnold, and lo! it spake thus: I'm not exactly sure what you're all complaining about. The port just works fine, it does take a few seconds for make to iterate over all options, but it's because the port has way too many options. It's a pretty sizable difference... % svn co -r394572 [ Mk, print] % cd print/ghostscript9 % time make -V MAINTAINER doc...@freebsd.org 0.903u 0.023s 0:00.92 100.0%834+264k 82+0io 6pf+0w % ( cd ../../Mk svn up -r394573 ) Updating '.': Ubsd.options.mk Ubsd.port.mk Updated to revision 394573. % time make -V MAINTAINER doc...@freebsd.org 81.908u 0.070s 1:21.99 99.9%848+264k 82+0io 5pf+0w On my E3-1230v2 system (not a slouch), the difference is 1s to 110s - which is in line with Matthew's results. A two-orders-of-magnitude slowdown in make startup is unreasonable and I formally request r394573 be backed out. I can see the benefits of the functionality but the current cost is too high. -- Peter Jeremy pgpl7TlEjyf80.pgp Description: PGP signature
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
On Thu, 20 Aug 2015, Henk van Oers wrote: [root@mbox /usr/ports/print/ghostscript9]# time date Thu Aug 20 13:49:43 CEST 2015 real0m0.003s user0m0.003s sys 0m0.001s [root@mbox /usr/ports/print/ghostscript9]# time make clean [...] [root@mbox /usr/ports/print/ghostscript9]# time date Thu Aug 20 15:01:46 CEST 2015 real0m0.004s user0m0.003s sys 0m0.000s Err, why did you feel the need to time the date command? And as root? -- Dave Horsfall DTM (VK2KFU) Those who don't understand security will suffer. RIP Cecil the Lion; he was in pain for two days, thanks to some brave dentist. ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
+--On 20 août 2015 08:24:25 +0100 Anton Shterenlikht me...@bris.ac.uk wrote: | OTOH I do not want to use ghostscript ! | | In my view ghost is a critically important port. | I will not be able to use FreeBSD productively withouth it. | Since I have no time or skill to contribute to | address the slowness problem, I'm happy to put | up with the extra hour. I'm not exactly sure what you're all complaining about. The port just works fine, it does take a few seconds for make to iterate over all options, but it's because the port has way too many options. It's not as if you're all sitting behind your desk looking at the screen waiting for it to finish. -- Mathieu Arnold pgp5AsVu4ZruT.pgp Description: PGP signature
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
+--On 19 août 2015 11:51:07 +0200 Luca Pizzamiglio luca.pizzamig...@gmail.com wrote: | Hi Henk, | same here. | | make -V PKGNAME in ghostscript9 takes a huge amount of time. | BTW, every make in ghostscript* takes a lot of time. | | I've no idea why. Because it stupidly has a gazillions options. Someone should reduce that to 3 or 4. -- Mathieu Arnold pgp4j8yQVabfY.pgp Description: PGP signature
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
+--On 19 aot 2015 11:51:07 +0200 Luca Pizzamiglio luca.pizzamiglio at gmail.com wrote: | Hi Henk, | same here. | | make -V PKGNAME in ghostscript9 takes a huge amount of time. | BTW, every make in ghostscript* takes a lot of time. | | I've no idea why. Because it stupidly has a gazillions options. Someone should reduce that to 3 or 4. I blame some stupid /usr/ports/Mk change. Last week it took ''portmaster -a'' less then a minute to say what to do. Now it takes an hour or so... :( OTOH I do not want to use ghostscript ! I'll reconfig ImageMagick-6.9.1.10_1,1 That's the one that pulled it in. Thx, Henk ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
Hi Henk, same here. make -V PKGNAME in ghostscript9 takes a huge amount of time. BTW, every make in ghostscript* takes a lot of time. I've no idea why. Best regards, Luca On Wed, Aug 19, 2015 at 5:45 AM, Henk van Oers h...@signature.nl wrote: All my ports are up to date, but running ''portmaster -va'' on ~1000 ports takes more then 50 minutes. Most time is spend on ghostscript9-9.06_10 I started a ''portmaster ghostscript9-9.06_10'' a while ago and it does not do any thing yet but using 100% CPU on one prosessor. What is going on? -- Henk ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Running ''portmaster ghostscript9-9.06_10'' takes a lot of time
All my ports are up to date, but running ''portmaster -va'' on ~1000 ports takes more then 50 minutes. Most time is spend on ghostscript9-9.06_10 I started a ''portmaster ghostscript9-9.06_10'' a while ago and it does not do any thing yet but using 100% CPU on one prosessor. What is going on? -- Henk ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
