Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On Mon, 17 Aug 2020 01:13:19 +0100
Ken Moffat via blfs-support
wrote:
> On Sun, Aug 16, 2020 at 05:28:42PM -0400, Scott Andrews via
> blfs-support wrote:
> > On Sun, 16 Aug 2020 22:10:37 +0100 Ken Moffat via blfs-support
> > wrote:
> >
> > > >
> > > For a desktop machine, 2GB is woefully inadequate nowadays :-(
> > > (And if it is a server, nss might be an uncommon choice).
> > >
> > > ĸen
> >
> > Maybe from your point a view. I am using raspbery pi 3 and 4 as
> > desktop machines and servers. They work well.
> >
> Great, for whatever packages you are compiling on your desktop. But
> the OP was using x86_64 and on that 2GB is woefully inadequate for
> comiling a modern desktop - unless someone is using a single core
> machine, and I do not recall when those were last produced in
> x86_64. To be honest I could not contemplate building and
> maintaining an x86_64 BLFS machine with less than 4 cores and 2GB
> RAM per core (for things like rust and qtwebengine). Oh, my laptop
> has 8 cores but a pitiful amount of RAM (6.7GB according to top,
> some is used for the video) and on that I have to shut down almost
> everything and force rust and qtwebengine to use fewer cores when I
> rebuild those or firefox, and it is painful.
>
> For i686 I have no idea how much RAM is needed to build and maintain
> a BLFS *desktop* system.
>
> I assume you are using 32-bit for at least pi3. Out of interest,
> Which graphical browser do you build for your desktop ?
>
> ĸen
Chromium and firefox-esr.
The build machine is a raspberry pi 2
A 900MHz quad-core ARM Cortex-A7 CPU
1GB RAM
100 Base Ethernet
4 USB ports
40 GPIO pins
Full HDMI port
Combined 3.5mm audio jack and composite video
Camera interface (CSI)
Display interface (DSI)
Micro SD card slot
VideoCore IV 3D graphics core
FLAGS are as follows
# ARM settings, RPI2
MARCH="armv7-a+neon-vfpv4"
MTUNE="cortex-a7"
MFLOAT="hard"
MFPU="neon-vfpv4"
%_optflags -O2 -pipe -march=${MARCH} -mtune=${MTUNE}
-mfloat-abi=${MFLOAT} -mfpu=${MFPU} -fPIC -fomit-frame-pointer
-ftree-vectorize
%_system_type --with-arch=${MARCH} --with-fpu=${MFPU}
--with-float=${MFLOAT} --with-arch-directory=arm
I can build LFS-9.0 complete with kernel, dovecot, exim, dns and rpm
in less than 24 hours.
I don't have the time for Chromium and firefox-esr, as I didn't keep
the logs for the old build system. I am currently updating my build
scripts to make the simpler.
--
http://lists.linuxfromscratch.org/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On Sun, Aug 16, 2020 at 05:28:42PM -0400, Scott Andrews via blfs-support wrote: > On Sun, 16 Aug 2020 22:10:37 +0100 Ken Moffat via blfs-support > wrote: > > > > > > For a desktop machine, 2GB is woefully inadequate nowadays :-( > > (And if it is a server, nss might be an uncommon choice). > > > > ĸen > > Maybe from your point a view. I am using raspbery pi 3 and 4 as > desktop machines and servers. They work well. > Great, for whatever packages you are compiling on your desktop. But the OP was using x86_64 and on that 2GB is woefully inadequate for comiling a modern desktop - unless someone is using a single core machine, and I do not recall when those were last produced in x86_64. To be honest I could not contemplate building and maintaining an x86_64 BLFS machine with less than 4 cores and 2GB RAM per core (for things like rust and qtwebengine). Oh, my laptop has 8 cores but a pitiful amount of RAM (6.7GB according to top, some is used for the video) and on that I have to shut down almost everything and force rust and qtwebengine to use fewer cores when I rebuild those or firefox, and it is painful. For i686 I have no idea how much RAM is needed to build and maintain a BLFS *desktop* system. I assume you are using 32-bit for at least pi3. Out of interest, Which graphical browser do you build for your desktop ? ĸen -- Juliet's version of cleanliness was next to godliness, which was to say it was erratic, past all understanding and was seldom seen. -- Unseen Academicals -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On Sun, 16 Aug 2020 22:10:37 +0100 Ken Moffat via blfs-support wrote: > > > For a desktop machine, 2GB is woefully inadequate nowadays :-( > (And if it is a server, nss might be an uncommon choice). > > ĸen Maybe from your point a view. I am using raspbery pi 3 and 4 as desktop machines and servers. They work well. -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On Sun, Aug 16, 2020 at 11:25:54AM -0500, Douglas R. Reno via blfs-support wrote: > > On 8/16/20 10:58 AM, Hans Malissa via blfs-support wrote: > > On August 16, 2020 at 6:27 AM, Xi Ruoyao via blfs-support > > wrote: > > > On 2020-08-16 02:27 +0100, Ken Moffat via blfs-support: > > > > Playing with CFLAGS does not always do what you expect (it depends > > > > on the individual packages as to whether you need to take special > > > > action to force your own CFLAGS, and trying to detune released > > > > packages seems like a bad idea. For the little it is worth, I did > > > > some experiments just over a year ago with the aim of forcing my own > > > > CFLAGS, CXXFLAGS and exploring some of the options. The results > > > > (basically one run of each variation, but with some upgrades along > > > > the way) were mostly inconclusive but somewhere in there are details > > > > of what I had to do to the packages I build to get them to obey my > > > > CFLAGS (or in some cases, to not use my optimization of -O2 or -O3) > > > > because some default to -O3 but will detune to -O2 if you pass that, > > > > and one some of my less-powerful machines I do generally use -O2. > > > > > > > > But the problem was in nss. I do not regard that as a large > > > > package, although it is a slow one when built using -j1. > > > > AFAICS building nss-3.55 less than 300 MB which should be trivial. > > > > > > Current version of NSS can be built with -jN. But I can tell that > > > the test > > > suite just fails with -O3. > > > > I followed the build with free from another terminal; the machine does > > indeed run out of memory (2GB on this system) during make. > > I temporarily added some swap space (on an external hard drive), and the > > build succeeded - building NSS needs ~1GB of swap on top of my 2GB. > > I'm planning to install NSS and remove the swap space again. Does the > > memory usage during the build have anything to do with memory usage > > during run-time? Will I be able to use NSS without the swap space? > > By the way, the test suite (following the instructions on > > http://linuxfromscratch.org/blfs/view/systemd/postlfs/nss.html ) > > reports: > > > > Tests summary: > > -- > > Passed: 54674 > > Failed: 2442 > > Failed with core: 0 > > ASan failures: 0 > > Unknown status: 8 > > TinderboxPrint:Unknown: 8 > > > > Is it safe to proceed with 2442 failed tests? > > Greetings, > > > > Hans > > > You should be safe to proceed. The NSS test suite is quite buggy, and it's > dependent on the contents of /etc/hosts as well as you passing > "HOST=localhost DOMSUF=localdomain". On that note, make sure that you have > the following line in /etc/hosts next time you run the tests: > > 127.0.0.1 localhost.localdomain localhost > > I think that was changed after the 9.1 release cycle. That allows the NSS > test suite to function properly. > > You should be OK, NSS should work perfectly without swap space. I think you > hit the memory ceiling when linking. You might want to invest in more RAM > though in the future, if at all possible. > > > - Doug > For a desktop machine, 2GB is woefully inadequate nowadays :-( (And if it is a server, nss might be an uncommon choice). ĸen -- Juliet's version of cleanliness was next to godliness, which was to say it was erratic, past all understanding and was seldom seen. -- Unseen Academicals -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] Beyond Linux® From Scratch (System V Edition) - Version 9.1: LSB-Tools-0.6
On Sun, 16 Aug 2020 18:25:12 + (UTC) DJ Lucas via blfs-support wrote: > On 8/16/2020 7:15 AM, Scott Andrews via blfs-support wrote: > > On Fri, 14 Aug 2020 12:53:39 -0400 Scott Andrews > > wrote: I have patched and extended > > and changed the name of lsbtools: It make lsbtools a functional > > utility for keeping systemv init scripts in the proper order. I > > merged the remove_initd and install_initd into a single pythons > > script and added s "service and display function. I changed the > > name to chkconfig chkconfig returns: S 0 1 2 3 4 5 6 checkfs S30 > > --- --- --- --- --- --- --- cleanfs S45 --- --- --- --- --- --- --- > > console S70 --- --- --- --- --- --- --- halt --- S99 --- --- --- > > --- --- --- localnet S08 S90 --- --- --- --- --- S90 modules S05 > > --- --- --- --- --- --- --- mountfs S40 S70 --- --- --- --- --- S70 > > mountvirtfs S00 --- --- --- --- --- --- --- network --- K80 K80 K80 > > S20 S20 S20 K80 rc --- --- --- --- --- --- --- --- reboot --- --- > > --- --- --- --- --- S99 sendsignals --- S60 --- --- --- --- --- S60 > > setclock --- --- --- --- --- --- --- --- swap S20 S65 --- --- --- > > --- --- S65 sysctl S90 --- --- --- --- --- --- --- sysklogd --- K90 > > K90 K90 S10 S10 S10 K90 template --- --- --- --- --- --- --- --- > > udev S10 --- --- --- --- --- --- --- udev_retry S50 --- --- --- --- > > --- --- --- chkconfig --add
Re: [blfs-support] Beyond Linux® From Scratch (System V Edition) - Version 9.1: LSB-Tools-0.6
On 8/16/2020 7:15 AM, Scott Andrews via blfs-support wrote: On Fri, 14 Aug 2020 12:53:39 -0400 Scott Andrews wrote: I have patched and extended and changed the name of lsbtools: It make lsbtools a functional utility for keeping systemv init scripts in the proper order. I merged the remove_initd and install_initd into a single pythons script and added s "service and display function. I changed the name to chkconfig chkconfig returns: S 0 1 2 3 4 5 6 checkfs S30 --- --- --- --- --- --- --- cleanfs S45 --- --- --- --- --- --- --- console S70 --- --- --- --- --- --- --- halt --- S99 --- --- --- --- --- --- localnet S08 S90 --- --- --- --- --- S90 modules S05 --- --- --- --- --- --- --- mountfs S40 S70 --- --- --- --- --- S70 mountvirtfs S00 --- --- --- --- --- --- --- network --- K80 K80 K80 S20 S20 S20 K80 rc --- --- --- --- --- --- --- --- reboot --- --- --- --- --- --- --- S99 sendsignals --- S60 --- --- --- --- --- S60 setclock --- --- --- --- --- --- --- --- swap S20 S65 --- --- --- --- --- S65 sysctl S90 --- --- --- --- --- --- --- sysklogd --- K90 K90 K90 S10 S10 S10 K90 template --- --- --- --- --- --- --- --- udev S10 --- --- --- --- --- --- --- udev_retry S50 --- --- --- --- --- --- --- chkconfig --add
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On 8/16/20 10:58 AM, Hans Malissa via blfs-support wrote: On August 16, 2020 at 6:27 AM, Xi Ruoyao via blfs-support wrote: On 2020-08-16 02:27 +0100, Ken Moffat via blfs-support: Playing with CFLAGS does not always do what you expect (it depends on the individual packages as to whether you need to take special action to force your own CFLAGS, and trying to detune released packages seems like a bad idea. For the little it is worth, I did some experiments just over a year ago with the aim of forcing my own CFLAGS, CXXFLAGS and exploring some of the options. The results (basically one run of each variation, but with some upgrades along the way) were mostly inconclusive but somewhere in there are details of what I had to do to the packages I build to get them to obey my CFLAGS (or in some cases, to not use my optimization of -O2 or -O3) because some default to -O3 but will detune to -O2 if you pass that, and one some of my less-powerful machines I do generally use -O2. But the problem was in nss. I do not regard that as a large package, although it is a slow one when built using -j1. AFAICS building nss-3.55 less than 300 MB which should be trivial. Current version of NSS can be built with -jN. But I can tell that the test suite just fails with -O3. I followed the build with free from another terminal; the machine does indeed run out of memory (2GB on this system) during make. I temporarily added some swap space (on an external hard drive), and the build succeeded - building NSS needs ~1GB of swap on top of my 2GB. I'm planning to install NSS and remove the swap space again. Does the memory usage during the build have anything to do with memory usage during run-time? Will I be able to use NSS without the swap space? By the way, the test suite (following the instructions on http://linuxfromscratch.org/blfs/view/systemd/postlfs/nss.html ) reports: Tests summary: -- Passed: 54674 Failed: 2442 Failed with core: 0 ASan failures: 0 Unknown status: 8 TinderboxPrint:Unknown: 8 Is it safe to proceed with 2442 failed tests? Greetings, Hans You should be safe to proceed. The NSS test suite is quite buggy, and it's dependent on the contents of /etc/hosts as well as you passing "HOST=localhost DOMSUF=localdomain". On that note, make sure that you have the following line in /etc/hosts next time you run the tests: 127.0.0.1 localhost.localdomain localhost I think that was changed after the 9.1 release cycle. That allows the NSS test suite to function properly. You should be OK, NSS should work perfectly without swap space. I think you hit the memory ceiling when linking. You might want to invest in more RAM though in the future, if at all possible. - Doug -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On August 16, 2020 at 6:27 AM, Xi Ruoyao via blfs-support wrote: On 2020-08-16 02:27 +0100, Ken Moffat via blfs-support: Playing with CFLAGS does not always do what you expect (it depends on the individual packages as to whether you need to take special action to force your own CFLAGS, and trying to detune released packages seems like a bad idea. For the little it is worth, I did some experiments just over a year ago with the aim of forcing my own CFLAGS, CXXFLAGS and exploring some of the options. The results (basically one run of each variation, but with some upgrades along the way) were mostly inconclusive but somewhere in there are details of what I had to do to the packages I build to get them to obey my CFLAGS (or in some cases, to not use my optimization of -O2 or -O3) because some default to -O3 but will detune to -O2 if you pass that, and one some of my less-powerful machines I do generally use -O2. But the problem was in nss. I do not regard that as a large package, although it is a slow one when built using -j1. AFAICS building nss-3.55 less than 300 MB which should be trivial. Current version of NSS can be built with -jN. But I can tell that the test suite just fails with -O3. I followed the build with free from another terminal; the machine does indeed run out of memory (2GB on this system) during make. I temporarily added some swap space (on an external hard drive), and the build succeeded - building NSS needs ~1GB of swap on top of my 2GB. I'm planning to install NSS and remove the swap space again. Does the memory usage during the build have anything to do with memory usage during run-time? Will I be able to use NSS without the swap space? By the way, the test suite (following the instructions on http://linuxfromscratch.org/blfs/view/systemd/postlfs/nss.html ) reports: Tests summary: -- Passed: 54674 Failed: 2442 Failed with core: 0 ASan failures: 0 Unknown status: 8 TinderboxPrint:Unknown: 8 Is it safe to proceed with 2442 failed tests? Greetings, Hans -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] 9.1-systemd: NSS-3.55 build fails
On 2020-08-16 02:27 +0100, Ken Moffat via blfs-support: > Playing with CFLAGS does not always do what you expect (it depends > on the individual packages as to whether you need to take special > action to force your own CFLAGS, and trying to detune released > packages seems like a bad idea. For the little it is worth, I did > some experiments just over a year ago with the aim of forcing my own > CFLAGS, CXXFLAGS and exploring some of the options. The results > (basically one run of each variation, but with some upgrades along > the way) were mostly inconclusive but somewhere in there are details > of what I had to do to the packages I build to get them to obey my > CFLAGS (or in some cases, to not use my optimization of -O2 or -O3) > because some default to -O3 but will detune to -O2 if you pass that, > and one some of my less-powerful machines I do generally use -O2. > > But the problem was in nss. I do not regard that as a large > package, although it is a slow one when built using -j1. > AFAICS building nss-3.55 less than 300 MB which should be trivial. Current version of NSS can be built with -jN. But I can tell that the test suite just fails with -O3. -- Xi Ruoyao School of Aerospace Science and Technology, Xidian University -- http://lists.linuxfromscratch.org/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-support] Beyond Linux® From Scratch (System V Edition) - Version 9.1: LSB-Tools-0.6
On Fri, 14 Aug 2020 12:53:39 -0400 Scott Andrews wrote: I have patched and extended and changed the name of lsbtools: It make lsbtools a functional utility for keeping systemv init scripts in the proper order. I merged the remove_initd and install_initd into a single pythons script and added s "service and display function. I changed the name to chkconfig chkconfig returns: S 0 1 2 3 4 5 6 checkfs S30 --- --- --- --- --- --- --- cleanfs S45 --- --- --- --- --- --- --- console S70 --- --- --- --- --- --- --- halt--- S99 --- --- --- --- --- --- localnetS08 S90 --- --- --- --- --- S90 modules S05 --- --- --- --- --- --- --- mountfs S40 S70 --- --- --- --- --- S70 mountvirtfs S00 --- --- --- --- --- --- --- network --- K80 K80 K80 S20 S20 S20 K80 rc --- --- --- --- --- --- --- --- reboot --- --- --- --- --- --- --- S99 sendsignals --- S60 --- --- --- --- --- S60 setclock--- --- --- --- --- --- --- --- swapS20 S65 --- --- --- --- --- S65 sysctl S90 --- --- --- --- --- --- --- sysklogd--- K90 K90 K90 S10 S10 S10 K90 template--- --- --- --- --- --- --- --- udevS10 --- --- --- --- --- --- --- udev_retry S50 --- --- --- --- --- --- --- chkconfig --add
