Re: [gentoo-user] recovering pam.d directory
On August 8, 2012 04:22:28 PM Alan McKinnon wrote: > On Wed, 8 Aug 2012 13:24:17 + > > Marcello Varisco wrote: > > Hi all, > > > > By incident I removed the pam.d directory containing all pam modules > > from command line. Is there a way to recover the removed directory? > > any help is appreciated since I can't login to my computer without > > live cd anymore. > > Unless you have some amazing recovery tools to had (most people don't) > you can't easily recover those files. > > pam.d isn't put there by a single package, everything that uses pam is > liable to write it's own custom file there. You can regain your ability > to log in by remerging these packages: > > sys-apps/shadow > sys-auth/pambase > > To do that, you will need to boot off a livecd and chroot. Then a > reboot should see you fine. Then you could rememrge everything that has > pam in USE and hope this is enough. > > Or, you could restore from backups. You *do* have backups of /etc, > right? Also, qcheck from portage-utils can list out all packages that are missing files from /etc/pam.d, so you can know which packages need remerging. Run "qcheck --all" and look for "AFK: /etc/pam.d/." in the output. HTH, Bryan
Re: [gentoo-user] Where to discuss ARM stuff.
On 08/09/2012 07:24 AM, Norman Rieß wrote:
> Am 08.08.2012 14:14, schrieb Raffaele BELARDI:
> I have a problem compiling binutils.
> Full build log can be found here http://smash-net.org/temp/build_log.txt
>
> I tried tha vanilla flag and diabled zlib as suggested via google
> search, but neither did help.
> I also tried latest unstable version 2.29 i think.
Quite obscure...
I'd try to reproduce by issuing only the offending line from the shell
(the one that compiles "options.cc", see below) and then try again
removing the -O2 flag.
Also check if the disk is full ("end of file not at end of a line;
newline inserted" is suspicious but it could just be a side effect of
the compiler crash).
There are some build requirements listed in the README under
/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/,
double check if they are met by your ARM toolchain.
If all fails you could try disabling "gold" compilation (the
--enable-gold bit) and just compile/use the standard GCC linker.
Sorry, not much else I can think of. I never had to compile the binutils
package because I had a binary cross-compiler toolchain for ARM available.
raf
$ armv6j-hardfloat-linux-gnueabi-g++ -DHAVE_CONFIG_H -I.
-I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold
-I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold
-I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/../include
-I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/../elfcpp
-DLOCALEDIR="\"/usr/share/binutils-data/armv6j-hardfloat-linux-gnueabi/2.22/locale\""
-DBINDIR="\"/usr/armv6j-hardfloat-linux-gnueabi/binutils-bin/2.22\""
-DTOOLBINDIR="\"/usr/armv6j-hardfloat-linux-gnueabi/bin\"" -W -Wall
-D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -frandom-seed=options.o -O2
-pipe -mcpu=arm1176jzf-s -mfpu=vfp -MT options.o -MD -MP -MF
.deps/options.Tpo -c -o options.o
/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/options.cc
Re: [gentoo-user] Want to seriously test a NEW hard drive
On Wed, 08 Aug 2012 22:22:25 -0500, Dale wrote: > I used something called Gparted to partition this monster. This is > temporary tho. What should I be using to partition this thing? What is > the best, and easiest, tool for me to use? I have been using cfdisk in > the past but it doesn't seem to work on this one. Do not use cfdisk on 2TB+ drives, it gives bad alignment. Use gdisk or cgdisk with GPT partition tables. Create a 1MB partition at the start of type EF02, which keeps backward compatibility, then partition as normal. Note that GPT has none of the primary/logical crap, just create the partitions you want. -- Neil Bothwick Downloading - A quick way of catching a virus from anywhere in the world. signature.asc Description: PGP signature
Re: [gentoo-user] Want to seriously test a NEW hard drive
Neil Bothwick wrote: > On Wed, 08 Aug 2012 22:22:25 -0500, Dale wrote: > >> I used something called Gparted to partition this monster. This is >> temporary tho. What should I be using to partition this thing? What is >> the best, and easiest, tool for me to use? I have been using cfdisk in >> the past but it doesn't seem to work on this one. > Do not use cfdisk on 2TB+ drives, it gives bad alignment. Use gdisk or > cgdisk with GPT partition tables. Create a 1MB partition at the start of > type EF02, which keeps backward compatibility, then partition as normal. > > Note that GPT has none of the primary/logical crap, just create the > partitions you want. > > A, so that 1Mb partition is what does the alignment thingy? Is it the same on all large drives or does it vary a bit based on size? Thanks. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Want to seriously test a NEW hard drive
On Thu, 09 Aug 2012 04:35:15 -0500, Dale wrote: > > Do not use cfdisk on 2TB+ drives, it gives bad alignment. Use gdisk or > > cgdisk with GPT partition tables. Create a 1MB partition at the start > > of type EF02, which keeps backward compatibility, then partition as > > normal. > > > > Note that GPT has none of the primary/logical crap, just create the > > partitions you want. > > > > > > A, so that 1Mb partition is what does the alignment thingy? Is it > the same on all large drives or does it vary a bit based on size? No, gdisk/cgdisk does the alignment. The 1MB GPT boot partition give backward compatibility with DOS booting. The size doesn't vary, I've used the same on drives from 120GB to 3TB. -- Neil Bothwick the sum of all human intelligence is constant, only the number of humans increases. signature.asc Description: PGP signature
Re: [gentoo-user] Want to seriously test a NEW hard drive
Neil Bothwick wrote: > On Thu, 09 Aug 2012 04:35:15 -0500, Dale wrote: > >>> Do not use cfdisk on 2TB+ drives, it gives bad alignment. Use gdisk or >>> cgdisk with GPT partition tables. Create a 1MB partition at the start >>> of type EF02, which keeps backward compatibility, then partition as >>> normal. >>> >>> Note that GPT has none of the primary/logical crap, just create the >>> partitions you want. >>> >>> >> A, so that 1Mb partition is what does the alignment thingy? Is it >> the same on all large drives or does it vary a bit based on size? > No, gdisk/cgdisk does the alignment. The 1MB GPT boot partition give > backward compatibility with DOS booting. The size doesn't vary, I've used > the same on drives from 120GB to 3TB. > > Hmmm, then I guess I didn't do something right the first time I tried cgdisk then. It said it wasn't aligned right but I couldn't figure out how to get it to do it. Then I used Gparted and it seemed to do it without me telling it anything. The only difference I saw was the little 1Mb partition. Well, I plan to dd the thing and start over with LVM and all so I'll see what it does next time. Maybe I just didn't do something right. Since I have nothing windoze on here, do I really need to worry about DOS booting? All I have is Gentoo here. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Getting apvlv running (My "excellent adventure")
Walter Dnes [12-08-09 05:11]: > Earlier this year, my ISP changed their billing notification emails > from application/pdf to application/octet-stream. Trying to view it > from mutt showed binary gobbledygook. After some flailing around, I > found out that I had to put an entry into .mailcap, namely... > > application/octet-stream; mimeopen %s > > When I tried to open a pdf file from mutt, I got a text dialogue > asking me which program to use. I chose /usr/bin/epdfview, and mutt has > used that as the default ever since. > > Now epdfview is masked for removal in a few weeks. apvlv is the > recommended lightweight alternative. After some screwing around and > discovering an obscure bug in the apvlv ebuild, I finally got apvlv up > and running. You ***MUST*** build poppler with USE="xpdf-headers", or > else the apvlv ebuild dies. I reported the bug, and the apvlv ebuild > now should check for app-text/poppler[xpf-headers]. > > I emerged and ran rox-mime-editor and have no clue what to do to > change from epdfview to apvlv. There are no man or info files for > rox-mime-editor. Is there a better alternative mime-editor? > > As a heavy-handed solution, I searched for the string "epdfview" in > ~/.local. In ~/.local/share/applications/defaults.list I found an entry > for pdf using epdfview. I zapped that line, and tried reading the pdf > in mutt. I got the text dialogue again, and specified /usr/bin/apvlv. > mutt now uses it all the time for pdf files. In contrast, Firefox is > much easier, with a dialogue for applications. > > -- > Walter Dnes > Hi Walter, thanks for the hints! :) Now apvlv works for me... Best regards, mcc
Re: [gentoo-user] Getting apvlv running (My "excellent adventure")
On 08/09/2012 02:06 PM, meino.cra...@gmx.de wrote: > Walter Dnes [12-08-09 05:11]: >> Now epdfview is masked for removal in a few weeks. apvlv is the >> recommended lightweight alternative. After some screwing around and >> discovering an obscure bug in the apvlv ebuild, I finally got apvlv up >> and running. You ***MUST*** build poppler with USE="xpdf-headers", or >> else the apvlv ebuild dies. I reported the bug, and the apvlv ebuild >> now should check for app-text/poppler[xpf-headers]. >> > > Hi Walter, > > thanks for the hints! :) > Now apvlv works for me... > Strange, I didn't have to add that USE for apvlv to build. Looks like it should be pulled in by the ebuild: $ grep xpdf-headers /usr/portage/app-text/poppler/poppler-0.20.2-r1.ebuild IUSE="cairo cjk curl cxx debug doc +introspection jpeg jpeg2k +lcms png qt4 tiff +utils +xpdf-headers" >From man 5 ebuild: IUSE This should be a list of any and all USE flags that are leveraged within your build script. ... Beginning with EAPI 1, it is possible to prefix flags with + or - in order to create default settings that respectively enable or disable the corresponding USE flags. But it did not work for you so probably there is a bug somewhere. raf
Re: [gentoo-user] Want to seriously test a NEW hard drive
Dale writes: > I have seen where people use dd to do this sort of thing to. I read > somewhere that if you do a dd and put in all 1's, then all 0's then back > again that it is very hard to get any data back off the drive. I think > if you do it like over a dozen times, it is deemed impossible to get > anything back. I think that is the Government standard of it's gone. There's no need for multiple passes of dd with different values. http://www.h-online.com/security/news/item/Secure-deletion-a-single-overwrite-will-do-it-739699.html Wonko
Re: [gentoo-user] Want to seriously test a NEW hard drive
Alex Schuster wrote: > Dale writes: > >> I have seen where people use dd to do this sort of thing to. I read >> somewhere that if you do a dd and put in all 1's, then all 0's then back >> again that it is very hard to get any data back off the drive. I think >> if you do it like over a dozen times, it is deemed impossible to get >> anything back. I think that is the Government standard of it's gone. > There's no need for multiple passes of dd with different values. > > http://www.h-online.com/security/news/item/Secure-deletion-a-single-overwrite-will-do-it-739699.html > > Wonko > > I wonder what some Government org like NSA would think about this? Then again, they may want us to believe this so they can get stuff back. ;-) ;-) That said, I always wondered how something can be there when it is erased. On paper, I can see that because it made a physical change to the paper but on magnetic media, it is magnetic not physical. Anyway. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Want to seriously test a NEW hard drive
On Thu, 09 Aug 2012 08:53:27 -0500 Dale wrote: > Alex Schuster wrote: > > Dale writes: > > > >> I have seen where people use dd to do this sort of thing to. I > >> read somewhere that if you do a dd and put in all 1's, then all > >> 0's then back again that it is very hard to get any data back off > >> the drive. I think if you do it like over a dozen times, it is > >> deemed impossible to get anything back. I think that is the > >> Government standard of it's gone. > > There's no need for multiple passes of dd with different values. > > > > http://www.h-online.com/security/news/item/Secure-deletion-a-single-overwrite-will-do-it-739699.html > > > > Wonko > > > > > > > I wonder what some Government org like NSA would think about this? > Then again, they may want us to believe this so they can get stuff > back. ;-) ;-) > > That said, I always wondered how something can be there when it is > erased. On paper, I can see that because it made a physical change to > the paper but on magnetic media, it is magnetic not physical. It's quite simple once you understand how disks work. In textbooks we need to keep things simple, so we say things like "the magnetic particles are all aligned this way for a 0 and that way for a 1". This gets the concept across but it also let's people believe that bits on a disk are very much binary - like a light switch or a transistor they are either on or off. In practice, nothing could be further from the truth. With disk magnetic media, you aren't dealing with a single isolated thing (such as a chunk of disk that can only be one way), you are dealing with a very very large number of magnetic particles that go to make up one bit. It's how they average out that makes the drive think it's a 0 or a 1. It all works much like tape cassettes - the head has a little coil of wire in it and current flows through the coil. It passes near a magnetic field on the tape, and the current in the coil changes. Read the amount of change using fancy electronics, and voila! you have audio. In the case of disks, it's voila! you have a stream of bits. Disk drives can't afford to be almost right like audio tape though, they have to be exactly right so the drive has some amazing maths built into it for error-correction and redundancy. I believe something like 40% of the space containing raw data is pure error checking (so your 3T drive is actually 4.1T, but you will only ever get to use 3T) The trick is, when you overwrite an area of the disk, you don't erase everything, there are some traces left behind. Pencil and paper is a good analogy. Write something in pencil. Erase wit with a rubber eraser, and write something else in the same space. Now hold the paper up to the light and if you know how to look you can see the indents in the paper from the first thing you wrote. Train your eyes to ignore what's written there now and only look at dented paper with no lead marks, and you can read things quite clearly. James Bond was especially good at this but that's a movie so real life isn't *that* quick. Sekrit magic disk software does a very similar thing - it ignores the current data and looks for traces left behind from the last write, and the ones before that. This trick isn't universal of course. As drive technology advances and IBM figures out new ways to do it, they come up with ideas like using the _depth_ of the magnetic material too. Neat trick - you can double the data stored in the same surface area. With each technology advance, things change a lot, so the amount of reading backwards you can do is always changing and depends very much on exactly what drive you have. -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] kernels & swap usage
http://people.gnome.org/~michael/blog/2012-08-08-libreoffice-3-6-0.html As someone said, a lot of legacy code is being removed/cleaned/simplyfied. Kernel could be an issue when it comes to swappiness, but not when it comes to the amount of ,memory that a given version of $CC will use to compile a given package, provided that $HARDWARE and $CFLAGS are constant.
Re: [gentoo-user] kernels & swap usage
120809 Jesús J. Guerrero Botella wrote: > http://people.gnome.org/~michael/blog/2012-08-08-libreoffice-3-6-0.html Yes, I read that c 2 hr ago (smile). > As someone said, a lot of legacy code is being removed/cleaned/simplified. It doesn't explain why I saw the same effect with Firefox 14.1 : perhaps there's been some code-cleaning there too. > Kernel could be an issue when it comes to swappiness, > but not when it comes to the amount of memory > a given version of $CC will use to compile a given package, > provided that $HARDWARE and $CFLAGS are constant. That's what I suspected, but wasn't sure about. Any other thoughts are welcome. -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
[gentoo-user] Anyone compiled libreoffice-3.6.0.4 yet?
This has been slow and painful so far. First, the build stops repeatedly because of zero-length library files. I can restart the ebuild manually and each iteration builds one more (real) library. I've been doing this iterating for hours and I think I may have gotten past that part, but the build is far from done... Second, the ebuild is using only one CPU out of four, so this is taking much longer than before. Ugh. Anyone else seeing these problems?
Re: [gentoo-user] Anyone compiled libreoffice-3.6.0.4 yet?
walt writes: > This has been slow and painful so far. > > First, the build stops repeatedly because of zero-length library files. > I can restart the ebuild manually and each iteration builds one more > (real) library. I've been doing this iterating for hours and I think > I may have gotten past that part, but the build is far from done... > > Second, the ebuild is using only one CPU out of four, so this is taking > much longer than before. Ugh. > > Anyone else seeing these problems? No, I just built it today, in the usual time of about 2.5 hours. Wonko
Re: [gentoo-user] Anyone compiled libreoffice-3.6.0.4 yet?
Am 09.08.2012 22:36, schrieb Alex Schuster: > No, I just built it today, in the usual time of about 2.5 hours. ~1 hr here :-P (I *had* to say that ;-) )
Re: [gentoo-user] Anyone compiled libreoffice-3.6.0.4 yet?
2012/8/9 Stefan G. Weichinger > Am 09.08.2012 22:36, schrieb Alex Schuster: > > > No, I just built it today, in the usual time of about 2.5 hours. > > ~1 hr here :-P > > (I *had* to say that ;-) ) > > > Hello, No problem building libreoffice-3.6.0.4. About 1h 38 min. Best regards -- Jacques
Re: [gentoo-user] Anyone compiled libreoffice-3.6.0.4 yet?
On Thu, Aug 9, 2012 at 3:19 PM, walt wrote: > This has been slow and painful so far. > > First, the build stops repeatedly because of zero-length library files. > I can restart the ebuild manually and each iteration builds one more > (real) library. I've been doing this iterating for hours and I think > I may have gotten past that part, but the build is far from done... > > Second, the ebuild is using only one CPU out of four, so this is taking > much longer than before. Ugh. > > Anyone else seeing these problems? No problem here, either. It needs a huge amount of temp disk space, be sure you're not running out (especially if it is on /dev/shm for example).
Re: [gentoo-user] Getting apvlv running (My "excellent adventure")
On Thu, Aug 09, 2012 at 03:07:38PM +0200, Raffaele BELARDI wrote > Strange, I didn't have to add that USE for apvlv to build. > Looks like it should be pulled in by the ebuild: > > $ grep xpdf-headers /usr/portage/app-text/poppler/poppler-0.20.2-r1.ebuild > IUSE="cairo cjk curl cxx debug doc +introspection jpeg jpeg2k +lcms png > qt4 tiff +utils +xpdf-headers" > > >From man 5 ebuild: > IUSE This should be a list of any and all USE flags that are leveraged > within your build script. ... Beginning with EAPI 1, it is possible > to prefix flags with + or - in order to create default settings that > respectively enable or disable the corresponding USE flags. > > But it did not work for you so probably there is a bug somewhere. "IUSE" is "Inherited USE". The "bug", if you want to call it that, is that I start my USE variable with "-*", which overrides default settings. The bugfix is to make app-text/poppler[xpdf-headers] mandatory for the apvlv ebuild. In that case, the ebuild will halt with an error message stating that app-text/poppler[xpdf-headers] is needed. This beats the ebuild compile dying with a message about missing files. -- Walter Dnes
Re: [gentoo-user] Where to discuss ARM stuff.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Am 09.08.2012 10:04, schrieb Raffaele BELARDI:
> On 08/09/2012 07:24 AM, Norman Rieß wrote:
>> Am 08.08.2012 14:14, schrieb Raffaele BELARDI: I have a problem
>> compiling binutils. Full build log can be found here
>> http://smash-net.org/temp/build_log.txt
>>
>> I tried tha vanilla flag and diabled zlib as suggested via
>> google search, but neither did help. I also tried latest unstable
>> version 2.29 i think.
>
> Quite obscure...
>
> I'd try to reproduce by issuing only the offending line from the
> shell (the one that compiles "options.cc", see below) and then try
> again removing the -O2 flag.
>
> Also check if the disk is full ("end of file not at end of a line;
> newline inserted" is suspicious but it could just be a side effect
> of the compiler crash).
>
> There are some build requirements listed in the README under
> /var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/,
>
>
double check if they are met by your ARM toolchain.
>
> If all fails you could try disabling "gold" compilation (the
> --enable-gold bit) and just compile/use the standard GCC linker.
>
> Sorry, not much else I can think of. I never had to compile the
> binutils package because I had a binary cross-compiler toolchain
> for ARM available.
>
> raf
>
> $ armv6j-hardfloat-linux-gnueabi-g++ -DHAVE_CONFIG_H -I.
> -I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold
>
>
- -I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold
> -I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/../include
>
>
-
-I/var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/../elfcpp
> -DLOCALEDIR="\"/usr/share/binutils-data/armv6j-hardfloat-linux-gnueabi/2.22/locale\""
>
>
- -DBINDIR="\"/usr/armv6j-hardfloat-linux-gnueabi/binutils-bin/2.22\""
> -DTOOLBINDIR="\"/usr/armv6j-hardfloat-linux-gnueabi/bin\"" -W
> -Wall -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64
> -frandom-seed=options.o -O2 -pipe -mcpu=arm1176jzf-s -mfpu=vfp -MT
> options.o -MD -MP -MF .deps/options.Tpo -c -o options.o
> /var/tmp/portage/sys-devel/binutils-2.22-r1/work/binutils-2.22/gold/options.cc
>
>
A guy from IRC #gentoo-embedded gave the essential hint yesterday.
Setting MAKEOPTS="-j1" instead of -j2 solves the problem.
Norman
-BEGIN PGP SIGNATURE-
Version: GnuPG v2.0.19 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iQEbBAEBAgAGBQJQJJAaAAoJEMCA6frkLT6zGmEH90zynjc862ofZ5S9vK6g8Srf
Ad9NjzrLXm9cYSfaTQAqJQYXEMr4fofjLQgk00wjW0Ofc7gHCY2BwjewLdN3KgIa
f+HVwd+KTxKTZ6kilVgQvZWB6ks/iRe6luJu1qzYDQ7n8QXCRRMGzUZ1jvFOVKOg
8v7xHs+dWppA/JnDMxu2lH2lb14D0U/9QIkvTMb7u/z5aXrRT3IMrDKdtpfhhsVj
AYhOjAQ6ii9j8SPOl/AlyAHxnJ0e1cmzNGCDEej2cZ4jAsbVvZAjtq660PBNsD+Y
5ZORsCTVOb3NL1z7UO493HS6D2NKmAFOGha6shvLtGlL6AACod5IVoO3MilU7g==
=BpwC
-END PGP SIGNATURE-
Re: [gentoo-user] kernels & swap usage
It could be anything. Maybe some orphaned process was running in the background and leaking ram, or something. It's futile to speculate now about that. Also, the -recently added- "pgo" USE flag could have something to do with that. Not sure, since I didn't bother to investigate it's true purpose on firefox. I really don't think that the kernel has changed in a significant way in this regard since the latests 2.6.x releases. But I certainly didn't read *all* the kernel changelogs. -- Jesús Guerrero Botella
