Hello Stuart Longland, (need to add the surname, there's a prominent Stuart (sthen@) around :-)
I have the gut feeling your are moving some Linux-isms to OpenBSD land. You are doing by far too much by yourself ;-) I need to add I've never used the loongson port. INSTALL.loongson talks of using the mips64el packages. (which currently have a timestamp of 2016-10-26) The easy way: 1) get http://yourmirr.openbsd.org/pub/OpenBSD/snapshots/loongson/bsd.rd 2) put bsd.rd to /bsd.rd 3) boot bsd.rd ('boot bsd.rd' at boot> prompt) 4) follow instructions 5) reboot 6) pkg_add <yourstuff> [or pkg_add -u if you already have the packages installed] (this usually "just works" because the installer puts the selected mirror into /etc/pkg.conf) 7) happy using until you need another package or you want to test a newer snapshot (which is a good thing for the project). stua...@longlandclan.id.au (Stuart Longland), 2016.11.16 (Wed) 10:19 (CET): > Hi Marcus, > On 16/11/16 18:48, Marcus MERIGHI wrote: > >> I initially installed Ports from the snapshot, but then encountered some > >> > 404 errors where package sources were no longer available. Thus, I > >> > figured I'd grab a version off the git mirror (github). > > Without a dmesg you are lowering your chances. > > Ahh okay, I omitted this as I saw this as a userland issue not a kernel > one, and dmesg can be rather long. It's the basic information they want and somewhat shows that you've read openbsd.org/mail.html and therefore are a good internet citizen ;-) > The dmesg in full: > > [ using 587952 bytes of bsd ELF symbol table ] > > Copyright (c) 1982, 1986, 1989, 1991, 1993 > > The Regents of the University of California. All rights reserved. > > Copyright (c) 1995-2016 OpenBSD. All rights reserved. > > http://www.OpenBSD.org > > > > OpenBSD 6.0 (GENERIC) #627: Thu Jul 28 21:15:50 MDT 2016 > > dera...@loongson.openbsd.org:/usr/src/sys/arch/loongson/compile/GENERIC This one's downloaded sunday 2016-11-13?! Currently the file sets on the servers have a date of 2016-11-15? > > real mem = 1073741824 (1024MB) > > avail mem = 1057538048 (1008MB) > > mainbus0 at root: Lemote Yeeloong > > cpu0 at mainbus0: STC Loongson2F CPU 797 MHz, STC Loongson2F FPU > > cpu0: cache L1-I 64KB D 64KB 4 way, L2 512KB 4 way > > bonito0 at mainbus0: memory and PCI-X controller, rev 1 > > pci0 at bonito0 bus 0 > > rl0 at pci0 dev 7 function 0 "Realtek 8139" rev 0x10: irq 5, address > > 00:23:8b:35:6f:71 > > rlphy0 at rl0 phy 0: RTL internal PHY > > smfb0 at pci0 dev 8 function 0 "Silicon Motion LynxEM+" rev 0xb0 > > wsdisplay0 at smfb0 mux 1: console (std, vt100 emulation) > > ohci0 at pci0 dev 9 function 0 "NEC USB" rev 0x44: irq 7, version 1.0 > > ehci0 at pci0 dev 9 function 1 "NEC USB" rev 0x05: irq 7 > > usb0 at ehci0: USB revision 2.0 > > uhub0 at usb0 "NEC EHCI root hub" rev 2.00/1.00 addr 1 > > glxpcib0 at pci0 dev 14 function 0 "AMD CS5536 ISA" rev 0x03: rev 3, 32-bit > > 3579545Hz timer, watchdog, gpio, i2c > > isa0 at glxpcib0 > > pckbc0 at isa0 port 0x60/5 irq 1 irq 12 > > pckbd0 at pckbc0 (kbd slot) > > wskbd0 at pckbd0: console keyboard, using wsdisplay0 > > pms0 at pckbc0 (aux slot) > > wsmouse0 at pms0 mux 0 > > mcclock0 at isa0 port 0x70/2: mc146818 or compatible > > ykbec0 at isa0 port 0x381/3 > > gpio1 at glxpcib0: 32 pins > > iic at glxpcib0 not configured > > glxclk0 at glxpcib0: clock, prof > > pciide0 at pci0 dev 14 function 2 "AMD CS5536 IDE" rev 0x01: DMA, channel 0 > > wired to compatibility, channel 1 wired to compatibility > > wd0 at pciide0 channel 0 drive 0: <WDC WD3200BEVT-11ZCT0> > > wd0: 16-sector PIO, LBA48, 305245MB, 625142448 sectors > > wd0(pciide0:0:0): using PIO mode 4, Ultra-DMA mode 2 > > pciide0: channel 1 ignored (disabled) > > auglx0 at pci0 dev 14 function 3 "AMD CS5536 Audio" rev 0x01: isa irq 9, > > CS5536 AC97 > > ac97: codec id 0x414c4760 (Avance Logic ALC655 rev 0) > > audio0 at auglx0 > > ohci1 at pci0 dev 14 function 4 "AMD CS5536 USB" rev 0x02: isa irq 11, > > version 1.0, legacy support > > ehci1 at pci0 dev 14 function 5 "AMD CS5536 USB" rev 0x02: isa irq 11 > > usb1 at ehci1: USB revision 2.0 > > uhub1 at usb1 "AMD EHCI root hub" rev 2.00/1.00 addr 1 > > usb2 at ohci0: USB revision 1.0 > > uhub2 at usb2 "NEC OHCI root hub" rev 1.00/1.00 addr 1 > > usb3 at ohci1: USB revision 1.0 > > uhub3 at usb3 "AMD OHCI root hub" rev 1.00/1.00 addr 1 > > apm0 at mainbus0 > > uvideo0 at uhub0 port 1 configuration 1 interface 0 "SuYin USB 2.0 Camera" > > rev 2.00/10.17 addr 2 > > video0 at uvideo0 > > umass0 at uhub1 port 1 configuration 1 interface 0 "Generic USB2.0-CRW" rev > > 2.00/58.87 addr 2 > > umass0: using SCSI over Bulk-Only > > scsibus0 at umass0: 2 targets, initiator 0 > > sd0 at scsibus0 targ 1 lun 0: <Generic-, Multi-Card, 1.00> SCSI0 0/direct > > removable serial.0bda0158114173400000 > > urtw0 at uhub1 port 4 configuration 1 interface 0 "Realtek RTL8187B" rev > > 2.00/2.00 addr 3 > > urtw0: RTL8187B rev E, address 00:17:c4:4d:ea:a0 > > vscsi0 at root > > scsibus1 at vscsi0: 256 targets > > softraid0 at root > > scsibus2 at softraid0: 256 targets > > pmon bootpath: /dev/disk/wd0 > > boot device: wd0 > > root on wd0a (43fc3fdb63b966d0.a) swap on wd0b dump on wd0b > > > I think I've understood you are running a snapshot as of some point in > > time? > > Yes, I basically downloaded a new snapshot on Sunday and unpacked that. > Sadly, I didn't keep it around to refer to. What exactly does "download and unpack" mean? Are you untarring stuff yourself? If yes, this won't get you over some API/ABI changes easily. Have you already had a look at openbsd.org/current.html? > > While you are on, say, a snapshot as of 2016-10-01 the project keeps > > moving. Depending on the point in the release process it takes hours to > > days until _your_ snapshot isn't the current snapshot anymore. > > > > packages(7) keep being built against the current snapshot, not _your_ > > snapshot. Library mismatches between *your* snapshot base system and the > > *projects* current ports happen soon. > > Ahh, so the ports tree is rolling-release and not tied to a particular > release of OpenBSD. This makes sense. Like the other trees (src, ports, xenocara, not www) releases are only cvs tags. snapshots too, I guess but have never seen much public information on the creation of snapshots apart from some posts on the mailing lists: Some times features get tested via snapshots. Then you get something via snapshots that's not yet in the real openbsd sources. But mostly it's just an arbitrary (yes, well, the name is appropriate) snapshot of what's currently in CVS. On faster architectures you get snapshots more often than on slower ones. Usually OpenBSD -current is supposed to compile and run. Exceptions confirm the rule ;-) Take a look at http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/ and open the pulldown menu. _BASE is what is shipped at release time. without "_BASE" is the release plus errata (openbsd.org/errata.html). > > But 404s? Have you checked whether the 404-files are on the servers with > > a different version number possibly? > > I'll have to go over more thoroughly which packages were causing the > download errors, when I saw them my first instinct was: "the ports are > out of date, try updating those first in case they've fixed the download > links". > > Upstream projects sometimes do move things around, and often without > telling people, so didn't see this as being an issue. > > If it continued to be a problem, I'd be noting what packages doing some > verification and possibly track down where the packages had moved to > before preparing a few patches. :-) > > So far after moving to the git version of the ports tree, that > particular issue has not returned, but I'm keeping an eye out. Ah I think we talk of two different things... Your download problems happened while building ports yourself, while doing "make fetch"? I was talking of "pkg_add foobar", which just downloads precompiled packages. > > Thus you need to install the wanted packages soon after putting a > > snapshot on your machine. > > > > (You could cvs check out the ports tree around the time you are > > downloading the snapshot and *hope* that this ports tree matches your > > snapshot base install and then build your packages yourself.) > > How quickly does it change? As mentioned I downloaded the snapshot on > Sunday and it was then or Monday evening, I noticed some packages were You downloaded it sunday 2016-11-20 but the kernel says it was compiled "Jul 28 21:15:50 MDT 2016". Darn old in amd64 land, don't know about loongson! > failing to download with missing upstream files. > > That's why I presumed my snapshot was possibly taken around the time of > the OpenBSD 6 release. (A wrong assumption.) > > The commands I followed were straight from the documentation: > https://www.openbsd.org/faq/faq15.html#PortsFetch Does https://www.openbsd.org/faq/faq15.html#Easy *not* work for you? What's the reason you compile stuff yourself? Out of personal interest: what does "uname -r -p" say on loongson? > >> > So I have a couple of packages (maybe about a dozen) installed via the > >> > snapshot ports and some via git ports. This might be the root of my > >> > problems. > > Are you talking about the openbsd-wip ports? > > If yes I think you are doing something unsupported and you won't get > > help from @openbsd people. > > > > While reading ports@ I often see that ports-people talk of things which > > have been made ready in openbsd-wip for inclusion in the ports tree. > > Right, so openbsd-wip is the "unstable" tree? I grabbed ports from > https://github.com/openbsd/ports/ which I understood to just be ports > CVS mirrored on git. Three reasons I went the git repo: > > 1. I know git well > 2. Using github means I'm not putting so much demand on the OpenBSD > project CVS servers There's CVS mirrors, too! > 3. If I find I need to make changes, I can branch it, and optionally, > export those changes out as patches to be submitted back (e.g. new > ports, fixes, etc) without needing to be on-line at the time. Branching is frowned upon. Send you (CVS) patches to ports@! (if you are close to current you chances are good) $ cd /usr/src $ vi bin/cat/cat.c $ cvs diff -u bin/cat/cat.c Index: bin/cat/cat.c =================================================================== RCS file: /cvs/src/bin/cat/cat.c,v retrieving revision 1.26 diff -u -p -u -r1.26 cat.c --- bin/cat/cat.c 19 Oct 2016 18:20:25 -0000 1.26 +++ bin/cat/cat.c 16 Nov 2016 10:49:08 -0000 @@ -247,3 +247,4 @@ raw_cat(int rfd) rval = 1; } } +/* testing cvs diff */ Find a decent mail client (not e.g. gmail) and send this inline. > If this is the -wip branch, then okay, I best get familiar with CVS and > get it (mainline branch) from the proper place. :-) Definitely if you want to compile yourself! for 6.0 stable: cvs -q -danon...@mirror.osn.de:/cvs update -P -rOPENBSD_6_0 -Pd for -current: alias cvsupd-cur="cvs -q -danon...@mirror.osn.de:/cvs update -PdA" cd /usr cvsupd-cur src cvsupd-cur ports cvsupd-cur xenocara (you sure can have them in other places, too; that's just the defaults...) cd ~/myopenbsd mkdir www cvsupd-cur www Now you don't have to access the interwebs to read the web pages. > >> > I now have seen the following a couple of times: > >>> > > ===> Building package for py-cairo-1.10.0p1 > >>> > > Create /usr/local/ports/packages/mips64el/all/py-cairo-1.10.0p1.tgz > >>> > > Error: Libraries in packing-lists in the ports tree > >>> > > and libraries from installed packages don't match > >>> > > --- /tmp/dep_cache.5mrOMvzTf/portstree-py-cairo-1.10.0p1 Wed > >>> > > Nov 16 09:11:03 2016 > >>> > > +++ /tmp/dep_cache.5mrOMvzTf/inst-py-cairo-1.10.0p1 Wed Nov 16 > >>> > > 09:11:04 2016 > >>> > > @@ -13,8 +13,8 @@ > >>> > > -W expat.11.0 > >>> > > -W fontconfig.10.0 > >>> > > -W freetype.25.0 > >>> > > --W glib-2.0.4200.4 > >>> > > --W gobject-2.0.4200.4 > >>> > > +-W glib-2.0.4200.3 > >>> > > +-W gobject-2.0.4200.3 > >>> > > -W m.9.0 > >>> > > -W pcre.3.0 > >>> > > -W pixman-1.32.6 > >>> > > *** Error 1 in /usr/ports/graphics/py-cairo > >>> > > (/usr/ports/infrastructure/mk/bsd.port.mk:3157 'wantlib-args') > >>> > > *** Error 1 in /usr/ports/graphics/py-cairo > >>> > > (/usr/ports/infrastructure/mk/bsd.port.mk:1888 > >>> > > '/usr/local/ports/packages/mips64el/all/py-cairo-1.10.0p1.tgz') > >> > > >> > I understand this is due to a discrepancy between the versions normally > >> > used in packages and those actually installed. I've been able to work > >> > around these using PKG_CREATE_NO_CHECKS=yes but I sense this is not a > >> > good idea long-term. > >> > > >> > What's the best way of fixing the underlying problem? Would manually > >> > updating (in this case) the glib and gobject ports fix it, or is there > >> > some other fix? > > I think you'd better remove all packages, put the latest snapshot on, > > reinstall the packages you want right away. > > Be a happy user after that until you 1) notice you need software from > > packages that are already ahead of your snapshot or 2) you have spare > > time to install the current snapshot of the day just for fun and testing > > (and increased likelyhood of help from @openbsd people, which are always > > talking -current). > > Ahh okay. I'll have a look into doing that. This isn't a production > machine, so if I balls it up, it's no big deal, the machine was > literally gathering dust. I personally (and from threads on misc@ I know others, too) run snapshots all the time on my notebook (lenovo x220), which is sort-of production use. I've been bitten only twice I seem to remember. > I don't mind re-loading, although compiling on this thing is a time > consuming process. And I still wonder why you insist on doing it ;-) Is it the "only self compiled is trustworthy" thing? Or does one have to compile the ports on loongson? Marcus > -- > Stuart Longland (aka Redhatter, VK4MSL) > > I haven't lost my mind... > ...it's backed up on a tape somewhere. > > > !DSPAM:582c24e580773851211028!
