Re: Bk6.6Ch6.9 glibc `__stack_chk_guard' work around.
Szabolcs Gyalókay wrote: > > gcc res_hconf.c -c -std=gnu99 -fgnu89-inline -O2 -Wall -Winline > -Wwrite-strings -fmerge-all-constants -g -Wstrict-prototypes > -DIS_IN_nscd=1 -D_FORTIFY_SOURCE=2 > -fpie > -fstack-protector > -DNOT_IN_libc=1 > -o /sources/glibc-build/nscd/res_hconf.o -MD -MP -MF > /sources/glibc-build/nscd/res_hconf.o.dt -MT Removing non-relevant stuff and reformatting a bit, my reference build has: gcc res_hconf.c -c -std=gnu99 -fgnu89-inline -O2 -Wall -Winline -Wwrite-strings -fmerge-all-constants -g -Wno-strict-prototypes -Wno-write-strings -Wstrict-prototypes -fexceptions -I../include -D_LIBC_REENTRANT -include ../include/libc-symbols.h -Dgethostbyname=res_gethostbyname -Dgethostbyname2=res_gethostbyname2 -Dgethostbyaddr=res_gethostbyaddr -Dgetnetbyname=res_getnetbyname -Dgetnetbyaddr=res_getnetbyaddr -o /sources/glibc-build/resolv/res_hconf.o -MD -MP -MF /sources/glibc-build/resolv/res_hconf.o.dt -MT /sources/glibc-build/resolv/res_hconf.o It would seem that your Makefile is significantly different from mine. That would indicate to me that there is something wrong with your Chapter 5 build as once you get into chroot, the builds really should be the same. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.4 Book
Baho Utot wrote: > I would like to download the LFS-6.4 version of the book but it is > missing from the download section > > http://archive.linuxfromscratch.org/lfs-museum/ > > Is it still available? Just a minor permissions problem. Both 6.3 and 6.4 are accessible now. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Invitation to connect on LinkedIn
Peeyush Chandel wrote: > LinkedIn User removed. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 32-bit or 64-bit Kernel prep and RAM
stosss wrote: > If I compile the kernel with 64GB instead of 4GB on a 32-bit system > that would allow me to use more than 4GB of physical RAM. Is my > understanding correct? Maybe. It depends on your hardware. If it's supported in HW, the kernel will do the equivalent of changing segments when it needs to address memory outside of the current 32-bit logical address. Of course this is not very efficient. > I have seen earlier that 64-bit does not improve performance very much > over 32-bit. My main interest in building a 64-bit system is to be > able to access more physical RAM. Will I be able to accomplish my > objective with the 64GB switch instead of the 4GB switch in the kernel > configuration? If you are using more than 4G at one time, you will see an improvement. Personally, I have 2G on my main system and it almost never swaps, so it wouldn't make much difference. OTOH, a large server may need it. For example: $ free total used free sharedbuffers cached Mem: 80684447991296 77148 0 404596364 -/+ buffers/cache: 33948924673552 Swap: 781249211907806621712 > Is there any other advantage to using a 64-bit system instead of a > 32-bit system? It really depends on your application. If you are doing a simulation of a complex fluid flow problem, it may be useful. If you are browsing the web, probably not. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 8.4.3. Testing the Configuration
Raj wrote: > on issuing the command grub> kernel /boot/grub/core.img i get the error > > unknown command kernel > > what should i do. It sounds like you already have GRUB2 installed. Try multiboot /boot/grub/core.img -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Kernel panic - not syncing: VFS; Unable to mount root fs on unknown-block(2.0)
Andrew Benton wrote:
> On 12/04/10 14:45, mas...@mail.com wrote:
>> Hi i got this error when i tried to boot in to my LFS system.
>> could it be that my fstab is screwed up, or is it something i forgot to
>> do with the kernel?
>
> It's definitely nothing to do with fstab, quite possibly something to do with
> your kernel config. I can boot this system with no /etc/fstab. It causes
> errors
> for the bootscripts but the kernel mounts the partition grub tells it to an it
> gets to a login prompt for root to login and fix the situation.
>
>> I'm using the LFS 6.6 book, and i am doing this in vmware.
vmware should be fine.
>> ### BEGIN /etc/grub.d/10_linux ###
>> menuentry "GNU/Linux, with Linux 2.6.32.8-lfs-6.6" {
>> insmod ext2
>> set root=(hd0,1)
>> search --no-floppy --fs-uuid --set
>> dd263110-8c28-48f1-bb0e-5c779c97e9ac
>> linux /vmlinux-2.6.32.8-lfs-6.6 root=/dev/sda2 ro
>> }
> Are these search lines needed? I don't put them in my grub.cfg
Andy, the search lines are added by GRUB, they are ignored by LFS, but I
agree that it would be clearer if they were removed. Otherwise, the
entry looks correct.
>> Kernel panic - not syncing: VFS; Unable to mount root fs on
>> unknown-block(2.0)
>> VFS: Cannot open root device "sda2" or unknown-block(2.0)
> My guess is that you haven't compiled support for your motherboard's chipset
> into
> your kernel. Use lspci to learn about your motherboard's chipset. Also, on
> your
> host system, use lsmod to see what modules it is using.
> It could be that you didn't compile libata and the kernel is calling
> the partition hda2. It could also be that you didn't compile in support for
> your
> root filesystem.
Not building a required driver into the kernel is almost certainly the
problem here. lsmod on an earlier system and lspci are exactly the
right tools to figure out the needed drivers. After that rebuilding the
kernel and retrying are the boot will be needed.
-- Bruce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: Gcc 4.5.0
Simon Geard wrote: > Anyone here had tried building LFS with the newly released gcc 4.5? Just > tried, using the same basic instructions as for current LFS (barring the > addition of MPC to GMP and MPFR), but get the following partway through > the 'make' step: > > checking for library containing strerror... configure: error: Link tests > are not allowed after GCC_NO_EXECUTABLES. > make[1]: *** [configure-target-libiberty] Error 1 > make[1]: Leaving directory `/mnt/lfs/sources/gcc-build' > make: *** [all] Error 2 > > I see a fair amount of history on gcc mailing lists with previous > versions, but no clear answers... Simon, read the last two days posts of -dev. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFSbook 6.6 Chap 6.9.1 Testsuite Results
Franz L. Kuhlmann wrote:
> (Sorry, I have to type manually - hence as few keystrokes as possible)
The format is really confusing, although I know what you are trying to do.
> Now, here is my question:
> Although tests failed which are NOT mentioned in the book, can I safely
> assume that these results will not frustrate my continuing work on LFS-6.6?
I can't say that I've seen those errors before. You don't give many
clues. What hardware are you building on? What is the host OS? Are
you do any customization at all? It's possible some of the errors could
be timing or host errors but to know, you need to look at the actual
output of the tests. For instance, the tst-cancel20.c test has several
places with things like:
if ( ... )
puts ("child thread: barrier_wait failed");
exit (1);
-- Bruce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: /usr/lib/libstdc++.so.6.0.14-gdb.py?
Andrew Benton wrote: > Hello all > I've recently started using gcc-4.5.0 and one of the features of it is that > it installs > a python script called libstdc++.so.6.0.14-gdb.py into /usr/lib. This causes > ldconfig > to complain every time it is run, so pretty much every one of my build logs > has a line > in it that looks like this: > > ldconfig: /usr/lib/libstdc++.so.6.0.14-gdb.py is not an ELF file - it has the > wrong magic bytes at the start. > > Does anyone know anything about this file? What's it for? Is it needed? I'm > tempted to > just remove it but I thought I'd try and find out a bit more about it first. It doesn't seem to hurt anything, but see http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41816 It looks like it was first identified in October. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: linux from scratch in 6.3 and section 5.7
stosss wrote: > Why are you building 6.3 when 6.6 is out and 6.7 will be coming out > some time in the near future. Not that near. We're trying 6 month schedule -- March and September. 6.6 should be good though. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFSbook 6.6 Chap 6.9.1 Testsuite Results
Franz L. Kuhlmann wrote: > Here are details extracted from "*glibc-check-log*": That's good info. What is the output of uname -a on your host in the the virtual system? It sounds like it may be an old kernel or one that is not compiled with needed options. The cannot open shared object file 'libgcc_s.so.1' probably can be fixed with a LD_LIBRARY_PATH definition, but that shouldn't be necessary. In any case, try make LD_LIBRARY_PATH=$(pwd)/$(../gcc-4.5.0/config.guess)/libgcc/ \ -k check 2>&1 | tee glibc-check-log Are you by any chance using gcc-4.5.0? -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.6 Section 6.22 typo?
duck wilson wrote: > Greetings, > > I am working thru LFS 6.6. In Section 6.22 (Coreutils-8.4) where running the > coreutils tests as user nobody the book gives the following command: > > su-tools nobody -s /bin/bash -c "make RUN_EXPENSIVE_TESTS=yes check" > > When I run this command, I cut-and-paste from the book, I get: > > root:/sources/coreutils-8.4# su-tools nobody -s /bin/bash -c "make > RUN_EXPENSIVE_TESTS=yes check" > bash: su-tools: command not found > > Should this command be: > > su -tools nobody -s /bin/bash -c "make RUN_EXPENSIVE_TESTS=yes check" > > That is, with a space after the su? Or maybe: > > su - tools nobody -s /bin/bash -c "make RUN_EXPENSIVE_TESTS=yes check" You either missed an instruction in section 5.17, Coreutils-8.4, cp -v src/su /tools/bin/su-tools or you don't have your PATH set up properly in chroot. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: gcc Pass 2 configure: error: unsupported system, cannot find sizeof (omp_lock_t)
JimD. wrote: > configure: error: unsupported system, cannot find sizeof (omp_lock_t) > make[1]: *** [configure-target-libgomp] Error 1 > make[1]: Leaving directory `/mnt/lfs/sources/gcc-build' > make: *** [all] Error 2 Are you using lfs-svn-20100420 or later? -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 5.9. Binutils-2.19.1 - Pass 2
Dmytro Boyko wrote: > Hi! > When I attempted execute "make" in section 5.9. Binutils-2.19.1 - Pass 2 > I got an error: > checking for off_t... yes > checking size of off_t... configure: error: in > `/mnt/LFS/source/binutils-build/bfd': > configure: error: cannot compute sizeof (off_t), 77 > See `config.log' for more details. > make[1]: *** [configure-bfd] Error 1 > > I don't have any idea what is this and what I should to do. > > P.S. LFS book ver. 6.5 1. I suggest LFS 6.6. 2. What is the host system? 3. Did you check the Host System Requirements? -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: disagrees about version of symbol module_layout
jumbophut wrote: > I'm not sure if it will fix the module issues, but the system probably > expects to find the System.map file at /boot/System.map-2.6.30.2-xxx, > where xxx is the value returned by uname -r on your system (it's 686 > on mine). System.map is only used by klogd. Kernel developers may need it, but it's doubtful that anyone else will need it. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
gaurav k wrote: > Hey Guys, > > This is my first mail to the list- I'm a relative newbie- though not a > complete Linux noob who is clueless about what is happening. > > > I am using Ubuntu 10.04- Lucid Lynx and I'm following LFS version 6.6. > > > My Host System Requirements output is as follows: > > bash, version 4.1.5(1)-release > /bin/sh -> /bin/dash Change to bash. > Binutils: (GNU Binutils for Ubuntu) 2.20.1-system.20100303 > bison (GNU Bison) 2.4.1 > /usr/bin/yacc -> /usr/bin/bison.yacc Is this a wrapper script pointing to GNU Bison? If so, it's OK. > bzip2, Version 1.0.5, 10-Dec-2007. > Coreutils: 7.4 > diff (GNU diffutils) 2.8.1 > find (GNU findutils) 4.4.2 > GNU Awk 3.1.6 > /usr/bin/awk -> /usr/bin/gawk > gcc (Ubuntu 4.4.3-4ubuntu5) 4.4.3 > GNU C Library (Ubuntu EGLIBC 2.11.1-0ubuntu7) stable release version 2.11.1 I don't know if this will cause a problem or not. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Invitation to connect on LinkedIn
Zhiyu Ren wrote: > LinkedIn User blocked. I sent a message to LinkedIn to try to get them to not send invitations to the linuxfromscratch.org domain. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: sysklogd 1.5.0 segfault,1.4.1 works well.
xinglp wrote: > I tested it on both LFS-SVN and LFS-6.5. Do I have to turn on > something in the kernel. You're a little short of info here. Does klogd run OK? Do you get any errors when you build? I don't know of anything different in the kernel, but which kernel are you using? Do you get any sysklog output at all, or does it segfault immediately? Does /etc/syslog.conf exist? What is your hardware? Are you running 32-bit or 64-bit? sysklogd 1.5.0 seems to run fine for me on a 64-bit system. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: sysklogd 1.5.0 segfault,1.4.1 works well.
xinglp wrote:
> strace result is like below
>
> [pid 440] read(2, "\n-8\n", 2048) = 11
> [pid 440] close(2)= 0
> [pid 440] munmap(0xb7723000, 4096)= 0
> [pid 440] write(1, "<6>May 5 21:58:10 kernel: klogd 1.5.0, log source =
> /proc/kmsg started.\0", 73) = 73
> [pid 440] uname({sys="Linux", node="3.4.lfs.6.5", ...}) = 0
> [pid 440] open("/boot/System.map-2.6.33.3", O_RDONLY) = -1 ENOENT (No such
> file or directory)
> [pid 440] open("/boot/System.map", O_RDONLY) = -1 ENOENT (No such file or
> directory)
> [pid 440] open("/System.map-2.6.33.3", O_RDONLY) = -1 ENOENT (No such file
> or directory)
> [pid 440] open("/System.map", O_RDONLY) = -1 ENOENT (No such file or
> directory)
> [pid 440] open("/usr/src/linux/System.map-2.6.33.3", O_RDONLY) = -1 ENOENT
> (No such file or directory)
> [pid 440] open("/usr/src/linux/System.map", O_RDONLY) = -1 ENOENT (No such
> file or directory)
> [pid 440] time(NULL) = 1273067890
> [pid 440] stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=127,
> ...}) = 0
> [pid 440] write(1, "<4>May 5 21:58:10 kernel: Cannot find map file.\0",
> 49) = 49
> [pid 440] open("/proc/kallsyms", O_RDONLY) = -1 ENOENT (No such file or
> directory)
> [pid 440] time(NULL) = 1273067890
> [pid 440] stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=127,
> ...}) = 0
> [pid 440] write(1, "<6>May 5 21:58:10 kernel: No module symbols loaded -
> kernel modules not enabled.\n\0", 83) = 83
> [pid 440] --- SIGSEGV (Segmentation fault) @ 0 (0) ---
It would be easy enough for you to copy th e System.map from your build
to "/boot/System.map-2.6.33.3" to avoid those messages. klogd is the
only thing that uses it.
System.map has not been really required before. It's generally only
been a one line message in the log. You might want to try enabling
kernel modules. I don't use many, but I think the kernel really wants
them to be enabled.
-- Bruce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: wget-list in svn-20100503
Philippe Delavalade wrote: > Hi. > > The wget-list in svn-20100503 tells to download zlib-1.2.4 but this is > nomore the current version and so, wget -i wget-list does not work > here. > > Is it possible to download version 1.2.5 or is it really safer to deal with > 1.2.4 ? 1.2.5 should be OK. We have it in our queue to update soon. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
Mike McCarty wrote: > Rather than try to figure out what packages actually require > bash features, and contact their maintainers, and then hope to > get fixes in place, they have chosen simply to say "make > /bin/sh point to a copy of bash". Actually, the last time I checked, one of the main packages that have the problem is glibc and it has been brought up. In the past they refused to change, although they only need to change #!/bin/sh to #!/bin/bash I don't know if this is the current status or not, however, there is really no advantage in not using bash. After all, it's the only shell available after you chroot. I really don't understand why some distros use dash. Yes, dash is smaller and faster, but although the differences are measurable, they are not discernible to the user. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
gaurav k wrote: > Hey Guys... > > I just thought I'd let you all know how it turned out. > > Chris, you were absolutely right. There was a package missing. I had > followed "host system requirements", but it seems that autoconf is > needed too. Ubuntu Lucid does not have it installed by default. If > you see the output I have pasted below- you'll see that autoconf was > missing. It is not mentioned in the "host system requirements". It > worked as soon as I installed it. I can certainly add autoconf to the host requirements, but the last time we checked, it was not a prerequsite for glibc. It should only be needed if configure.in is modified. We certainly build glibc in Chapter 6 without autoconf. > Simon- I really know just the basics about how bash works- but thank > you for pointing out that scripts might use /bin/sh- I had no clue! > But dash worked for me - luckily! In the past, dash did not work. It may be version dependent. I still think bash is the better option. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs-svn-9267 md5sum error
Andrew Benton wrote: > On 07/05/10 11:47, xinglp wrote: >> md5sum -c .\list.md5 >> >> lfs-bootscripts-20100124.tar.bz2: FAILED >> udev-config-20100128.tar.bz2: FAILED >> md5sum: WARNING: 2 of 76 computed checksums did NOT match >> > I wouldn't worry about those. The bootscripts and udev config change all the > time. > I think they're auto generated by a script every day so the timestamps on the > files > changes which gives a different md5sum? Nothing to worry about. That's right. For the md5sums to match, those files need to be downloaded on the same day that the book is generated (daily). It's only a problem for the svn versions of the book. The real key is the date embedded in the filenames. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: bounced mail to lfs-support list!?
Jannis Kafkoulas wrote: > I thought first my subscription wasn't valid any more and > therefore I tried to subscribe again. > > Does someone know what that all means? Not really. Your account looks good and you obviously got this through. I got a similar message last night and attributed to a spammer. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
GMP incorrectly detects CPU-VENDOR-OS triplet
Ticket #2661 On my x86_64 machine ./config.guess script of GMP 5.01 (and 5.00 as well) returns 'athlon-unknown-linux-gnu' which is incorrectly considered by ./configure script as 32-bit platform, so it expects sizeof(long) to be 4, when it is actually 8: checking compiler cc -O2 -pedantic -fomit-frame-pointer has sizeof (long)==4... no which results in a compiler error: configure: error: could not find a working compiler, see config.log for details This bug can be avoided by specifying --build=x86_64-unknown-linux-gnu option to ./configure script, but without knowing that your whole LFS assembly stops. Coreutils 8.4 uname -a Linux lfslivecd 2.6.22.5-64bit #1 SMP Thu Nov 29 07:28:23 GMT 2007 x86_64 x86_64 x86_64 GNU/Linux -- Looking at GMP. they have two scripts: config.guess and configfsf.guess. Their config.guess is returning athlon. You might want to try to run those to test the results. Looking at the contents of config.guess, it looks like the cpu is being misidentified. I'd try copying configfsf.guess to overwrite config.guess and see if that makes a difference. Also, what is the contents of `cat /proc/cpuinfo` -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: grub-install
Yan Mo wrote: > Hi, got past the previous issue by using a newer machine. > > Now have got upto section 8.4.2 in the LFS-BOOK-6.6 > > After entering: > > grub-install --grub-setup=/bin/true /dev/sda > > It gives the messages: > > grub-probe: error: cannot find a device for /boot/grub. > > No path or device is specified. > Try ``grub-probe --help'' for more information. > Auto-detection of a filesystem module failed. > Please specify the module with the options `--modules' explicitly. > > After doing a searchs of the Linux From Scratch website, I noticed that this > section of the book changes frequently. Should I try one of the previous > ways of doing this? I'm concerned I may not be able to reboot my computer if > I just try things... Did you exit/reenter chroot and forget to remount the special filesystems as specified in Section 6.2? You need to have at least /dev mounted and accessible while in chroot. I'm not sure about grub, but you may also need /sys and /proc. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: gzip help
Mag Gam wrote:
> I am trying to compile gzip using icc. I tried following these
> instructions,
> http://www.linuxfromscratch.org/hints/downloads/files/intel-c-compiler.txt
>
>
> But I keep getting,
>
> isnan.c(132): error: floating-point operation result is out of range
> static memory_double nan = { L_(0.0) / L_(0.0) };
> ^
>
> isnan.c(133): error: floating-point operation result is out of range
> static DOUBLE plus_inf = L_(1.0) / L_(0.0);
>
> Any tricks?
Wrong list. Google is your friend.
http://software.intel.com/en-us/forums/showthread.php?t=74369&p=1
-- Bruce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: cannot set root passwd
Mac Zero wrote: > I've just completed an jhalfs (latest development version) script of > LFS-SVN20100506. No errors were reported so I moved on to setting the root > password, editing /etc files, etc. > > I entered to the chroot using the command in section 6.4, but when I try and > set the root password I get this: > > root:/# passwd root > Changing password for root > Enter the new password (minimum of 5 characters) > Please use a combination of upper and lower case letters and numbers. > Bad password: too short. > Warning: weak password (enter it again to use it anyway). > passwd: password changed. > > The problem is I am not being prompted to enter a password. The command runs > and terminates immediately with the above output. The password appears to be > set (to what I do not know): It would appear that the keyboard input is not being picked. I'm not sure why, but check to see if /dev is populated correctly. If you are doing it from chroot after jhalfs, did you mount /dev into the chroot environment? See section 6.2. > root:/# cat /etc/passwd > > root:x:0:0:root:/root:/bin/bash > bin:x:1:1:bin:/dev/null:/bin/false > nobody:x:99:99:Unprivileged User:/dev/null:/bin/false Edit /etc/passwd to root::0:0:root:/root:/bin/bash ^^ And root can log on without a password until you get things straightened out. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc fails to make. (error 2)
William Immendorf wrote: > On Wed, May 19, 2010 at 5:36 PM, Kyle Brennan wrote: >> /usr/bin/install -c -m 644 >> /mnt/lfs/tools/glibc-build/../include/linux/limits.h >> /tools/include/linux/limits.h >> /usr/bin/install: `/mnt/lfs/tools/glibc-build/../include/linux/limits.h' and >> `/tools/include/linux/limits.h' are the same file > Wait, you're not supposed to compile your packges from the $LFS/tools > directory. Please, build in a seprate directory than your tools > directory. I agree with William. When you start Chapter 5, the only thing you should have in /mnt/lfs is tools and sources: /mnt/lfs/tools /mnt/lfs/sources Put all your sources in /mnt/lfs/sources and change to that directory. The directory name 'sources' is arbitrary, but it should not be tools and it needs to be in /mnt/lfs so the sources are available in Chapter 6. Then when you start, 5.4: tar -xf binutils-.tar.bz2 cd binutils- etc. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc make fails with Error 2
Kyle Brennan wrote: > Hi, > I am new here and I am attempting my first LFS build. I get to Glibc in > chapter 5.7 and everything goes smoothly until i get to the "make" > command. I run the command and it works for about fifteen minutes and > it Gives me the following error: > > make: *** [all] Error 2 > > then it quits to the prompt. That's a little light on info. Give us the last 20-30 lines before the final error message. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc make fails with Error 2
Kyle Brennan wrote: > /mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.4.3/../../../../i686-lfs-linux-gnu/bin/ld: > > cannot find -lgcc_eh It looks like a problem with the gcc build. You should have: /mnt/lfs/tools/gcc/x86_64-unknown-linux-gnu/4.4.3/libgcc_eh.a or /mnt/lfs/tools/gcc/i686-pc-linux-gnu/4.4.3/libgcc_eh.a -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc make fails with Error 2
Kyle Brennan wrote: > On 05/23/2010 06:07 PM, Bruce Dubbs wrote: >> Kyle Brennan wrote: >> >> >>> /mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.4.3/../../../../i686-lfs-linux-gnu/bin/ld: >>> >> >>> cannot find -lgcc_eh >>> >> It looks like a problem with the gcc build. You should have: >> >> /mnt/lfs/tools/gcc/x86_64-unknown-linux-gnu/4.4.3/libgcc_eh.a or >> /mnt/lfs/tools/gcc/i686-pc-linux-gnu/4.4.3/libgcc_eh.a >> >> -- Bruce >> >> > I got it from the tarball that was on the linux from scratch site. > there may be a problem with my host system. because when i was building > in the tools directory I did not reinstall. there may be a error there > because there is supposed to be a symlink there right? so if that is > correct, I would need to fix my host system. There wouldn't be a problem as long as you didn't install anything over a file in the /usr tree, but I suggest you delete /tools and any subdirectories in /mnt/lfs/sources and completely and start over. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc make fails with Error 2
Kyle Brennan wrote: >>> >> These types of errors usually means you don't have CFLAGS set correctly, >> which points to some problem with configparms. Do "cat configparms". >> > It worked, and just a quick question, when you go for pass 2, 3, etc... > do you start over with a clean build directory? > ex: > do you remove the binutils-build directory and then create a new one? > for the binutils pass 2 build? Yes! Remove both the binutils- directory and -build (if any) directory and re-extract from the tarballs. Do that for every package. Go back and re-read Chapter 5.3, Important Note 1. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc build CH6.9 Awk: command not found
Kyle Brennan wrote: > Hi again, > Big problem here, It appears that Gawk was improperly configured and now > when I try to configure my Glibc build, it runs into an error where it > cannot find the awk command. Read section vii. Host System Requirements. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glibc build CH6.9 Awk: command not found
Mike McCarty wrote: > Bruce Dubbs wrote: >> Kyle Brennan wrote: >>> Hi again, >>> Big problem here, It appears that Gawk was improperly configured and now >>> when I try to configure my Glibc build, it runs into an error where it >>> cannot find the awk command. >> Read section vii. Host System Requirements. > > In Ch 6.9 isn't the awk needed the one in .../tools? I missed that. We get so many posts that end up being because the Host Requirements aren't met that I thought this was another. Kyle, go back to 5.21. Gawk and reinstall. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: chapter 6.16: configure: error: in `/sources/gcc-build/i686-pc-linux-gnu/libgcc':, ... C preprocessor "/lib/cpp" fails sanity check
mhenriqu...@terra.cl wrote: > Hi Ken: > > Thanks for your reply and yeap, it seems that it was a issue related to > a non deleted build or source folder, I rebuild the toolchain from > chapter 5 and then again the toolchain of chapter 6 taking care of > delete those folders and now gcc final compilation pass, now running the > test, hopefully are going to be ok.. > > just to bother, the book mention something about that somewhere?, do I > miss something... 5.2 Important Note #1. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: > Have built LFS-4.1 & -6.1, trying to build -6.6 on -6.1. Stage 1 seemed > to go well. Building glibc early in Stage2 failed. I did a search, saw > others had a similar problem, but didn't see a relevant solution. The > /scripts in my $PATH is where I put my cut&pasted "build scripts". > Suggestions appreciated. (I can supply my log.conf if relevant.) > gcc res_hconf.c -c -std=gnu99 -fgnu89-inline -O2 -O3 -Wall -Winline ... > /usr/local/src/glibc-build/nscd/nscd.o: In function `nscd_open_socket': > /usr/local/src/glibc-2.11.1/nscd/nscd.c:442: undefined reference to > `__stack_chk_guard' We've seen this before and there was really no good answer. I did some googling and I think this may be a kernel issue. In the kernel .config I am using, there is # CONFIG_CC_STACKPROTECTOR is not set I'm using an X86_64 and not a i686, so that may make a difference. I found this reference: http://fxr.watson.org/fxr/ident?v=xnu-1456.1.26;im=bigexcerpts;i=__stack_chk_guard If your kernel was built with this parameter, you might want to try building under another kernel that does not have this set. Another thing to try might be to change the CFLAGS line to add -fno-stack-protector -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: chapter 6.16: configure: error: in `/sources/gcc-build/i686-pc-linux-gnu/libgcc':, ... C preprocessor "/lib/cpp" fails sanity check
Neal Murphy wrote: > On Wednesday 26 May 2010 08:33:20 linux fan wrote: >> On 5/26/10, Simon Geard wrote: >>> I can't really see how things can be improved (short of blinking red >>> text on *every* page), but it does seem like half the problems we deal >>> with are from people not following it. Are people not even reading that >>> entire "General Compilation Instructions" page, and skipping straight to >>> the packages? >> The important note #1 in ch 5.3 is clear. >> However, "General Compilation Instructions" never specifies the flow >> of events expected to follow: >> >> 1 - make sure $LFS is set to the target directory (export LFS=/mnt/lfs) >> 2 - change directory to $LFS/sources >> 3 - unpack the package tarball >> 4 - change directory to the unpacked directory created by tar >> 5 - follow package build and install instructions >> 6 - change directory to $LFS/sources >> 7 - remove the unpacked directory created by tar We don't have it in a list, but it is in Section 5.3. I even put in a note as the very first thing in Binutils: Note Go back and re-read the notes in the previous section. Understanding the notes labeled important will save you a lot of problems later. I guess I'll need to reemphasize some more. :( -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: >> We've seen this before and there was really no good answer. >> >> I did some googling and I think this may be a kernel issue. In the >> kernel .config I am using, there is >> >> # CONFIG_CC_STACKPROTECTOR is not set >> >> I'm using an X86_64 and not a i686, so that may make a difference. >> >> If your kernel was built with this parameter, you might want to try >> building under another kernel that does not have this set. > > When I installed LFS-6.1, it was patched up to 2.6.17, one short. ;-( > To > do this 6.6 install I patched to .18. Grep didn't find STACKPROTECTOR > in my .config file. I ran a menuconfig and looked in the "likely > places", but found no references to a stack protector option. Only > found a few lines grepping the whole kernel branch for stack_protector. > > If it were a hardware thing, wouldn't glibc's configure figure that out? > >> Another thing to try might be to change the CFLAGS line to add -fno-stack- >> protector > > That failed at the same place. This attempt wasn't able to override > its makefile parms, as shown here when it entered this directory. > > As this is glibc, I'm not anxious to "get creative". Don't know what to > do now. Perhaps search kernel.org for more info? The reference you > provided didn't initially look informative. I'll look again, I guess. > > make[2]: Entering directory `/usr/local/src/glibc-2.11.1/nscd' make[2]: > Leaving directory `/usr/local/src/glibc-2.11.1/nscd' make[2]: Entering > directory `/usr/local/src/glibc-2.11.1/nscd' > > gcc nscd.c -c -std=gnu99 -fgnu89-inline -O2 -O3 -Wall -Winline -Wwrite- > strings -fmerge-all-constants -fno-stack-protector -g < - > march=i486 -mtune=native -pipe -Wstrict-prototypes -mpreferred-stack- > boundary=2 -DIS_IN_nscd=1 -D_FORTIFY_SOURCE=2 -fpie -fstack-protector <<-- overrides earlier setting > -I../include -I/usr/local/src/glibc-build/nscd < > -I/usr/local/src/glibc- > build -I../sysdeps/i386/elf -I../nptl/sysdeps/unix/sysv/linux/i386/i686 > -I../sysdeps/unix/sysv/linux/i386/i686 - > I../nptl/sysdeps/unix/sysv/linux/i386 -I../sysdeps/unix/sysv/linux/i386 > -I../nptl/sysdeps/unix/sysv/linux -I../nptl/sysdeps/pthread - > I../sysdeps/pthread -I../sysdeps/unix/sysv/linux -I../sysdeps/gnu - > I../sysdeps/unix/common -I../sysdeps/unix/mman -I../sysdeps/unix/inet > -I../sysdeps/unix/sysv/i386 -I../nptl/sysdeps/unix/sysv - > I../sysdeps/unix/sysv -I../sysdeps/unix/i386 -I../nptl/sysdeps/unix - > I../sysdeps/unix -I../sysdeps/posix -I../sysdeps/i386/i686/fpu - > I../nptl/sysdeps/i386/i686 -I../sysdeps/i386/i686 -I../sysdeps/i386/i486 > -I../nptl/sysdeps/i386/i486 -I../sysdeps/i386/fpu -I../nptl/sysdeps/i386 > -I../sysdeps/i386 -I../sysdeps/wordsize-32 -I../sysdeps/ieee754/ldbl-96 > -I../sysdeps/ieee754/dbl-64 -I../sysdeps/ieee754/flt-32 - > I../sysdeps/ieee754 -I../sysdeps/generic/elf -I../sysdeps/generic - > I../nptl -I.. -I../libio -I. -D_LIBC_REENTRANT -include > ../include/libc- > symbols.h -DNOT_IN_libc=1 -o /usr/local/src/glibc-build/nscd/nscd.o -MD > -MP -MF /usr/local/src/glibc-build/nscd/nscd.o.dt -MT > /usr/local/src/glibc- > build/nscd/nscd.o After you get to this stoppage, try doing a paste of just these instructions back with modifications. To do that, you need to paste into vim and add backslashes to each line, change what you want (delete -fstack-protector) and see if you can get that one file (build/nscd/nscd.o) to build. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: > OK, I did that. It paused about one > second, then returned a command prompt. No console messages at all. > Ummm, waita minute! (I don't compile anything unless I'm installing > a package--I finished being a programmer over 25 years ago.) I guess > that means it worked, eh? Yes. > So now go back to my build script and > pick it up again with make and see if it will complete? Yes. > Meaning my glibc will have this one stack busting flaw forever? No, not a flaw, but without a feature. It means that there is a mismatch in your system between the kernel and the (upstream) glibc build system. Something is being detected when it shouldn't be (or vice versa). Since you are using a fairly old kernel, the problem is really yours. The only way we would be able to get anyone to pay attention is is we were using a recent system to do the build. On my reference build I have gcc nscd.c -c -std=gnu99 -fgnu89-inline -O2 -Wall -Winline -Wwrite-strings -fmerge-all-constants -g -Wstrict-prototypes -DIS_IN_nscd=1 -D_FORTIFY_SOURCE=2 -fpie -fstack-protector -I../include -I/sources/glibc-build/nscd -I/sources/glibc-build -I../sysdeps/x86_64/elf -I../nptl/sysdeps/unix/sysv/linux/x86_64 -I../sysdeps/unix/sysv/linux/x86_64 -I../sysdeps/unix/sysv/linux/wordsize-64 -I../nptl/sysdeps/unix/sysv/linux -I../nptl/sysdeps/pthread -I../sysdeps/pthread -I../sysdeps/unix/sysv/linux -I../sysdeps/gnu -I../sysdeps/unix/common -I../sysdeps/unix/mman -I../sysdeps/unix/inet -I../nptl/sysdeps/unix/sysv -I../sysdeps/unix/sysv -I../sysdeps/unix/x86_64 -I../nptl/sysdeps/unix -I../sysdeps/unix -I../sysdeps/posix -I../sysdeps/x86_64/fpu -I../nptl/sysdeps/x86_64 -I../sysdeps/x86_64 -I../sysdeps/wordsize-64 -I../sysdeps/ieee754/ldbl-96 -I../sysdeps/ieee754/dbl-64 -I../sysdeps/ieee754/flt-32 -I../sysdeps/ieee754 -I../sysdeps/generic/elf -I../sysdeps/generic -I../nptl -I.. -I../libio -I. -D_LIBC_REENTRANT -include ../include/libc-symbols.h -DNOT_IN_libc=1-o /sources/glibc-build/nscd/nscd.o -MD -MP -MF /sources/glibc-build/nscd/nscd.o.dt -MT /sources/glibc-build/nscd/nscd.o Without error. As I said earlier, I don't know why this problem appears to be triggered on some systems and not others. Another suggestion would be to download jhalfs and build LFS using that. http://www.linuxfromscratch.org/alfs/. Failure at the same point that you are getting now would rule out errors earlier in the build. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: >> The __stack_chk_guard which is only seen by grep in gcc, must have >> gotten built in at some earlier step or possibly bled in from the > > How far back do you suppose? Stage 2 pass 1 or Stage 1 pass 2? I can > quite confidently rip either out and replace it with my scripts. > (UPMPMG; use package management, package management good) > > I don't suppose you have any idea what kind of parameter error might > cause gcc's make to include it? /tools/lib does have libssp. I suppose > that's indicative, eh? (Can't give you a listing at the moment, I'm on > a different twinned box also running 6.1. But I can query my package > management on the other box to see which step added libssp.) In my reference build I have: -rw-r--r-- 1 root root 26098 Apr 15 18:09 /mnt/lfs/tools/lib/libssp.a -rwxr-xr-x 1 root root 925 Apr 15 16:37 /mnt/lfs/tools/lib/libssp.la lrwxrwxrwx 1 root root15 Apr 15 16:37 /mnt/lfs/tools/lib/libssp.so -> libssp.so.0.0.0 lrwxrwxrwx 1 root root15 Apr 15 16:37 /mnt/lfs/tools/lib/libssp.so.0 -> libssp.so.0.0.0 -rwxr-xr-x 1 root root 12721 Apr 15 18:09 /mnt/lfs/tools/lib/libssp.so.0.0.0 The contenst of libssp.la are: $ cat /mnt/lfs/tools/lib/libssp.la # libssp.la - a libtool library file # Generated by libtool (GNU libtool 1.3134 2009-11-29) 2.2.7a # # Please DO NOT delete this file! # It is necessary for linking the library. # The name that we can dlopen(3). dlname='libssp.so.0' # Names of this library. library_names='libssp.so.0.0.0 libssp.so.0 libssp.so' # The name of the static archive. old_library='libssp.a' # Linker flags that can not go in dependency_libs. inherited_linker_flags='' # Libraries that this one depends upon. dependency_libs='' # Names of additional weak libraries provided by this library weak_library_names='' # Version information for libssp. current=0 age=0 revision=0 # Is this an already installed library? installed=yes # Should we warn about portability when linking against -modules? shouldnotlink=no # Files to dlopen/dlpreopen dlopen='' dlpreopen='' # Directory that this library needs to be installed in: libdir='/tools/lib' --- $ nm -a /mnt/lfs/tools/lib/libssp.so.0.0.0|grep stack 1000 T __stack_chk_fail 1040 t __stack_chk_fail_local 00201d60 B __stack_chk_guard B is the uninitialized data section (known as BSS). -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: > So it seems the HSR needs GCC>4.2, kernel>2.6.19? I'm not sure about that, but it's possible. I have an older LFS system that I use for day-to-day work. It is an i686 originally built in Nov 2005 using gcc-4.0.2, Linux 2.6.12, glibc-2.3.6. (SVN-20051118, approximately LFS-6.1.1). I upgraded it to Linux-2.6.22.5 without problem (2007), but haven't upgraded tool change components since. I use another system for my LFS development work now. >> What if you built a Linux-2.6.22.5 kernel on the host and try that? > > I'd fear that too large a kernel upgrade would require other confuration > changes in my host system. A patch from what was 17, now 18, to 19 > would probably reduce those chances. No, I think 2.6.22.5 would be fine. Besides, the only thing you do is copy bzImage and boot to that. I don't think it would interfere with anything currently running, but if it did, the only thing needed to revert is a reboot. > So it's beginning to look like from my LFS-6.1 system, with gcc-3.4.3 > and linux-2.6.11.12, originally, patched to 2.6.18.0, is just too far > a chasm? Possibly. The problem may be not using gcc-4.x. > The 6.6 book's HSR needs patching? I need an interim step, > say my own build of 6.3? Is that where we are? If that's confirmed, > I'll abandon this 6.6 build until I've done that--but I'd like to have > all your best advice on that--it's twice as much work! My advice would be to use jhalfs for some arbitrary version of the book that you choose and just let it run. We know that LFS-6.3 works because it has been used a lot from the last LiveCD. That would be a good candidate for an intermediate release. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: >> My advice would be to use jhalfs for some arbitrary version of the >> book that you choose and just let it run. We know that LFS-6.3 works >> because it has been used a lot from the last LiveCD. That would be a >> good candidate for an intermediate release. > > No, if I decide to "double my fun", I'll make 6.3 a "normal" upgrade > from my current 6.1 version. I am determined to have a consistent path. > That's fine. But your book's HSR, that little script that's included, > tells me, except for the kernel version, my 6.1 system is "good to go." The Host System requirements may indeed be too low for LFS 6.6, but I am reluctant to change them based on your input because you have made lots of changes. You claim your scripts encompass the book's commands, but I don't have the time or desire to check that, especially when we have an automated way to build. If you want to help, fine. Give us definitive reasons to change the book. That includes validation of your findings using our tools. You don't need to change what you have. It's easy enough to clone directories and to do testing with that. > What's the HSR if not exactly that? Granted, it seems there wasn't > enough exploration done to verify what gcc/kernel was required, as > linuxfan recently posted. You seem to be demanding perfection from a small group of volunteers. We don't claim to be perfect. There are too many combinations to check old versions of the packages against the latest version. > Now that you have information that gcc-4.2 is required to compile the > stack protector kernel code, Where did that come from? I don't see a reference. > I expect at least an erratum to the HSR is called for. You don't have any standing to 'expect' anything from us. You can suggest, but with your attitude, my reaction is to push back and say no, even if that's wrong. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
x2...@lycos.com wrote: That was a very entertaining post. Thanks. On a more serious note, we do try to address problems that come up. The issue is that we have seen lots of posts where the final resolution is "oops, it was my error, not the book's". (Which you humorously noted.) The reason we ask for a "standard" build is so we can try to duplicate the problem. For instance, I cannot build the 'standard' -dev book right now because the kernel panics when built with the default configuration on x86_64 hardware. See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44129 I don't get much response from the gcc team, but I understand why. I suspect they can't duplicate the problem, although there have been reports of the same problem by others. Unlike LFS, they are even paid to develop gcc (but not by me of course). There are fundamentally two general situations where we change LFS: when there is an upstream fix or when we can duplicate the problem. To do otherwise would cause even more problems. As far as nscd.c goes, the real problem is to determine why some systems appear to need -lssp and others don't. Then a fix the glibc build system can be made. Unfortunately, the only ones that can do that are those that can duplicate the problem. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Blfs X-Window (lfs scratch 6.3)
David Expósito wrote: > Hello > > I wonder if anyone has compiled lfs-scratch 6.3, the X-window environment > Section 23 of BLFS. X Window System Environment There is a separate list for blfs-support. Please use that for non-lfs issues. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: > As seems > indicated by the current situation, someone should adopt a QC role, and > have one system that trails, i.e. has exactly the package versions > specified in the HSR, and verifies that each version of LFS does in fact > install flawlessly with those prerequisites. Volunteers welcomed. > I began programming in 1966, FORTRAN II, on an IBM-1620 I started in 1965. I still have my FORTRAN (not II) manual. I also used an IBM-1620. The mass storage device was punched cards. Also IBM 7090/7094, IBM 360 Series 60, and CDC 6600. >> You don't have any standing to 'expect' anything from us. You can >> suggest, but with your attitude, my reaction is to push back and >> say no, even if that's wrong. > > Petulance? You need ego strokes for "being helpful"? Such is > "professionalism", eh? People report problems and you blow them off? > Why are you involved? No, not petulance, frustration. You report a problem doing things *your* way and we *did* try to help. We ask for your help in confirming the problem in a way we can duplicate and you say no. Who is being petulant here? > So what's your argument? "We'll put whatever we please in our book, > and whether it works for you or not is none of our concern?" Why > bother? My argument is that we will fix problems that we can confirm. I still don't see why you won't help us help you. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
>> The Host System requirements may indeed be too low for LFS 6.6 Updated to the packages in LFS-6.3, known to work for LFS-6.6 Added erratum to website. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Baho Utot wrote: > On 06/01/10 15:19, Bruce Dubbs wrote: >> Paul Rogers wrote: >> >>> As seems >>> indicated by the current situation, someone should adopt a QC role, and >>> have one system that trails, i.e. has exactly the package versions >>> specified in the HSR, and verifies that each version of LFS does in fact >>> install flawlessly with those prerequisites. >> Volunteers welcomed. > > I'll volunteer...I can mess up anything :) Excellent. > At this moment I have just about recovered my LFS 6.5/BLFS devel system, > from my LFS-6.3 build script screw up which, Ahem over wrote LFS-6.5 > with the files/package from LFS-6.3. Care to guess when the host system > puked? > I am going to try it again as I am a gluten for punishment. Can you please explain a little more. You are doing to start with a new 6.3 LFS system and build LFS-6.6? Please do report back any successes or problems. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Baho Utot wrote: > The system I am using to compile this stuff is an AMD Phenom(tm) II X4 > 810 Processor that has 8G DDR3 memory. I use that system so I can make > mistakes really fast ;) ie things can go sour quickly and at a high rate > of speed. You usually don't get a chance to see them let alone try to > stop it. Once you hit the key its all over. Not even enough > time to yell out a Oh Shhhi. I recommend: build-instruction 2>&1 | tee -a build.log > I build the system(s) on a removable SATA 500G drive. Then rsync it to > the resulting victim system. Be careful as the kernel drivers are probably different on the systems. > After I get that completed I am going to try using the LFS-6.3 with > BLFS-6.3-svn to build LFS 6.6. All of this would be using i686. > > I can try building LFS-6.6 after I get the LFS-6.3 up and running then I > can to the LFS-6.3 BLFS combo. > > I can currently build the LFS-6.3 tools, that part works > I am looking at why the scripts over wrote the host instead of going to > the chroot. I think that I was not really in the chroot when I ran the > chapter 06 build script. OK, thanks for the update. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: New LFS & business plan
Mike McCarty wrote: >> time -- 2.5 - 3 days (being generous here) > > I think you are being optimistic, for the first build, anyway. > I'd put in 5 days for getting the first build up and running. This all depends on experience. Actually I think there is a sweet spot. Too much experience and people want to customize the first time and run into problems. The best time for a very experienced LFSer is probably about 4 hours on a reasonably fast machine. For the first time, a newbie could take anywhere from weeks to infinity. 3 to 5 days are both reasonable estimates. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs-6.6 booting on my vmware!, but allays checking consistence of sda1
mhenriqu...@terra.cl wrote: > One minor issue that is bugging me, is that there is something wrong > about the sda1 and ext3 filesystem about the dates, since each time that > I turn on the virtual machine the system stop at booting telling me that > the check date on the filesystem is in the future and it have to stop > the booting, so, next boot I enter into the Vmware BIOS and change the > date there to something in the future (let say 5-6-2011), then booting > my lfs it check that the cheking date expire,and it do a checkdisk over > the partition, after that it boots ok and all goes well. But on next > boot it again claim that the filesystem check date is in the future and > I have to do all over again... > > So, I wonder if there is some way to check what dates the ext3 > filesystem is taking into account to see what is wrong there or more > simple, a way to turn off that checking consistence of the filesystem > acording to a date on boot? It sounds like a mismatch between the vmware hwclock and the system clock. Generally it's easiest to use GMT for the hw clock. Make sure /etc/sysconfig/clock has UTC=1 and check the old, but still mostly valid, hint at http://www.linuxfromscratch.org/hints/downloads/files/time.txt -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs-6.6 booting on my vmware!, but allays checking consistence of sda1
mhenriqu...@terra.cl wrote: > On 06/02/2010 11:32 AM, Ken Moffat wrote: >> On 2 June 2010 16:05, mhenriqu...@terra.cl wrote: >>> So, I wonder if there is some way to check what dates the ext3 >>> filesystem is taking into account to see what is wrong there or more >>> simple, a way to turn off that checking consistence of the filesystem >>> acording to a date on boot? You can look at the fs superblock with dumpe2fs -h You can change several things with tune2fs. That may help you do what you want. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Mike McCarty wrote: > linux fan wrote: >> On 6/2/10, Danny Engelbarts wrote: >> >>> ... a 6.3 system is required than the book should state 6.3 until proven >>> otherwise. >> That is just exactly what the DEV book now requires. >> >> Unfortunately, now we don't get to find out exactly why the original problem >> "undefined reference to __stack_chk_guard" >> surfaces from time to time. > > This may be a better discussion of the "problem", if such it be. > > http://www.ecos.sourceware.org/ml/libc-alpha/2006-04/msg9.html That is a general description of the problem. The more specific problem that we haven't solved is why some LFS systems appear to need -lssp when building ncsd within the chroot environment but most (at least reported to us) systems don't need it. The version of glibc within chroot certainly should support ssp but the two reports we have didn't seem to find it automatically. Others cannot duplicate the problem. If we had consistent errors on most builds, we could fix it. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: perl, gdbm and perl's obsequious help
Neal Murphy wrote: > Some time back (LFS 6.4), I discovered that perl's configure program can > poison the build; it is designed to be extremely helpful by ferreting out > features of the host system to support. Specifically in my case, because it > found libgdbm on the host system, it configured perl to include support for > gdbm; this causes the build to fail later because perl can't find libgdbm. > Alas, such obsequiousness hoses the LFS build process. > > The standard LFS configure options for perl do not prevent this from > happening. For review, the options are: > -des -Dprefix=/tools -Dstatic_ext='Data/Dumper Fcntl IO POSIX' > > After days of gnashing teeth and rending garments (my hair is too short to > pull out in tufts) and parsing perl's configure script, I developed the > correct incantation: > -des -Dprefix=/tools -Dstatic_ext="Data/Dumper Fcntl IO POSIX" \ > -Donlyextensions="Data/Dumper Fcntl IO POSIX" \ > -Dglincpth="/tools/include" -Dglibpth="/tools/lib" \ > -Dinc_version_list=none > > The four extra options are required _when building the toolchain_ to ensure > absolutely that perl's configure > - configures those extensions and ONLY those extensions, and > - cannot examine the host for features to support > These options should work regardless of what is installed on the host build > system. Thanks for the report Neal. I wonder if the extra commands are needed now that gdbm is built in Chapter 6 before Perl. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
linux fan wrote: > On 6/2/10, Bruce Dubbs wrote: > >> The version of glibc within chroot certainly should support ssp but the >> two reports we have didn't seem to find it automatically. Others cannot >> duplicate the problem. If we had consistent errors on most builds, we >> could fix it. > > Could there be slight variances in "how" one "enters chroot" that > could possibly affect this? I would't expect that to be likely. Me either. > Once in chroot, gcc and glibc have already been built (twice for gcc). > How could the conditions on the host system be still affecting the > outcome? Well we do mount /proc, /sys, /def, /dev/pts, and /dev/shm from the host system. The glibc build system might be probing something like /proc. Ken's method of first building the LFS target kernel in the host system and rebooting to that would avoid potential problems here. It's possible that something from the host gets pulled along as one of the packages is built in Chapter 5. However, the whole idea behind Chapter 5 -> chroot Chapter 6 is to avoid that. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Can't set $PS1 properly under Ubuntu 10.04 host
littlebat wrote: > Hi, > > I am learning: 4.4. Setting Up the Environment: > http://www.linuxfromscratch.org/lfs/view/6.6/chapter04/settingenvironment.html > . My host system is Ubuntu 10.04. > > I found it can't properly to set $PS1 for user "lfs" with the command > below provided by LFS6.6 book under Ubuntu 10.04 host system. cat > ~/.bash_profile << "EOF" exec env -i HOME=$HOME TERM=$TERM PS1='\u:\w\$ ' /bin/bash EOF > The reason is Ubuntu will setup PS1 in /etc/bash.bashrc, so I think > maybe we should set this value in ~/.bashrc ? No, that's not how it works. Upon login as the lfs user, bash executes /etc/profile $HOME/.bash_profile in that order. The 'exec env -i' portion ignores the environment that was set by /etc/profile. The second part sets up 3 and only 3 environment variables. bash itself sets up a few more, but not PS1. $HOME/.bashrc is not run by default in an initial login. Make sure you are doing an initial login. su - lfs The dash is important. Without it, $HOME/.bash_profile is not run and only $HOME/.bashrc is run. Other sub-shells that are run by the lfs user do use $HOME/.bashrc Take a look at the invocation portion of the bash man page and also at the env man page. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: perl, gdbm and perl's obsequious help
Neal Murphy wrote: > On Thursday 03 June 2010 14:45:04 Chris Staub wrote: >> On 06/03/2010 11:33 AM, Neal Murphy wrote: >>> On Thursday 03 June 2010 04:08:33 Simon Geard wrote: > IIRC, the problem became apparent when the toolchain perl tried to > run in the chroot jail during the final build (Ch. 6 equivalent) > before the final build of perl. It didn't matter what I did; > toolchain perl always built with support for libgdbm and would fail > because it couldn't find that lib in the chroot jail. It built in > support for libgdbm because it found gdbm libs and headers on my > Debian host. It found those things because the '-Dstatic_ext' option > does not prevent it from doing so. Hmm. I wonder why the incorrect behavior doesn't show up for me. On my current svn build, I do not have libgdbm in /tools, but I do have it in the final /usr/lib. ldd on perl does shows: # ldd perl linux-vdso.so.1 => (0x7fffb53ff000) libnsl.so.1 => /lib/libnsl.so.1 (0x7f912c312000) libdl.so.2 => /lib/libdl.so.2 (0x7f912c10e000) libm.so.6 => /lib/libm.so.6 (0x7f912be8c000) libcrypt.so.1 => /lib/libcrypt.so.1 (0x7f912bc55000) libutil.so.1 => /lib/libutil.so.1 (0x7f912ba52000) libc.so.6 => /lib/libc.so.6 (0x7f912b6f6000) /lib64/ld-linux-x86-64.so.2 (0x7f912c52a000) The tools version gives me: # ldd perl linux-vdso.so.1 => (0x7fffe93ff000) libnsl.so.1 => /tools/lib/libnsl.so.1 (0x7f342df99000) libdl.so.2 => /tools/lib/libdl.so.2 (0x7f342dd95000) libm.so.6 => /tools/lib/libm.so.6 (0x7f342db13000) libcrypt.so.1 => /tools/lib/libcrypt.so.1 (0x7f342d8dc000) libutil.so.1 => /tools/lib/libutil.so.1 (0x7f342d6d9000) libc.so.6 => /tools/lib/libc.so.6 (0x7f342d37d000) libgcc_s.so.1 => /tools/lib/libgcc_s.so.1 (0x7f342d168000) /tools/lib64/ld-linux-x86-64.so.2 (0x7f342e1b1000) I do have libgdm on the host system also. I didn't do the research that is in the book's dependencies section, but it says that perl is needed to build glibc. If it were a general problem, I'd think we would have seen it a long time ago. We may end up adding the extra switches in Chapter 5's perl build, but I'm trying to understand why it's needed on your system and not others before doing that. Figuring that out may be more work than anyone wants to do. > - This oddity has not been reported before (here or anywhere). - This > is not a problem report or support request. - It's a report of > something I encountered whilst using 6.4 as a guide to updating a > project that was originally based on LFS. - To the best of my ability > to determine, perl's configure script explicitly searches the host > filesystem for features it should support, an action which _can_ > poison the toolchain. - I know that the normal 'static_ext' parameter > does not prevent perl's configure from searching the host for > features to support. - I am very sure that the extra four configure > options do prevent perl's configure from potentially poisoning the > toolchain with references to the host ysstem. > > According to Ch. 5, the intent is to have perl build only the four > specified features, and build them statically into perl. As I > understand, the '-Dstatic_ext=...' option that LFS uses tells perl > only to build those features statically into perl. That's the intent, yes. > It doesn't > restrain perl from building other extensions. It doesn't prevent > perl's configure script from searching the host system. > > The real issue is that, as best as I can tell, perl's configure > script searches the host's filesystems for features to support, which > is not desired when building the toolchain. Why it does that to me > when building on Debian Lenny and not to anyone else, I don't know. I > do know that the extra parameters for perl's configure effectively > limit it to searching ONLY the toolchain for features to support. > > Perhaps I was remiss in not putting 'FYI' in the subject, No, not really. We appreciate and thank you for your report. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: >> shouldn't be a cross compiler. IE, it shouldn't have been configured >> with --target=$LFS_TGT. I wonder if the people reporting these errors >> mistakenly used the option --target=$LFS_TGT when they compiled gcc the >> second time in chapter 5? > > Yes, I can definitely confirm it was NOT in my Stage2 script. > > I got around the nscd error with: > > sed -i 's/fstack/fno-stack/' nscd/Makefile > > Then this evening I made a copy of the script without that line > and recompiled glibc with the Stage2 compiler as the next step. > No problem this time. So I think I have a fairly straightforward > workaround. I don't imagine there was any need for the nscd code > to get through the gcc compile. Now I think I have a good glibc > for the rest of the Stage2 build. Onward! So the problem was the Chapter 5 gcc? Let me review. I know you said your base host was LFS 6.1. Are you going directly to LFS 6.6 or an intermediate version? > I do have some errors to check out: > make[2]: *** [/usr/local/src/glibc-build/posix/tst-vfork3.out] Error 1 > make[2]: [/usr/local/src/glibc-build/posix/annexc.out] Error 1 (ignored) > make[1]: *** [posix/tests] Error 2 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-mutex8.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cond8.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-sem11.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-sem12.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancel24.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancelx4.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancelx5.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancelx16.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancelx20.out] Error 1 > make[2]: *** [/usr/local/src/glibc-build/nptl/tst-cancelx21.out] Error 1 > make[1]: *** [nptl/tests] Error 2 > make: *** [check] Error 2 Those seem fairly familiar to me for older systems. There may be some timing issues. annexc *always* fails. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
Paul Rogers wrote: >>> Then this evening I made a copy of the script without that line and >>> recompiled glibc with the Stage2 compiler as the next step. No >>> problem this time. So I think I have a fairly straightforward >>> workaround. I don't imagine there was any need for the nscd code to >>> get through the gcc compile. Now I think I have a good glibc for >>> the rest of the Stage2 build. Onward! >> So the problem was the Chapter 5 gcc? > > I'm not sure that follows. Mightn't it be something in the glibc > configure? It seems it had the opportunity to discover that stack > protector option would fail, but tried to compile it anyway. > >> Let me review. I know you said your base host was LFS 6.1. Are you >> going directly to LFS 6.6 or an intermediate version? > > Is there some good reason I shouldn't continue with 6.6 now? No. I was just asking for clarification. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: the issue of environment value $PS1 set under Ubuntu 10.04
littlebat wrote: >> Try adding "--noprofile --rc ~lfs/.bashrc" to the bash command. Or >> just > > Ubuntu bash has a bug, even if I specified --rcfile, it still read > /etc/bash.bashrc. That's because /etc/profile does it explicitly: PS1='\...@\h:\w\$ ' if [ -f /etc/bash.bashrc ]; then . /etc/bash.bashrc fi Read and understand the OPTIONS and INVOCATION sections of the bash man file. >> add "--noprofile --norc" and set everything in the env command. > > Base on my knowledge, I found the env command can't set like "set +h" > or "umask 022" those are set in the file .bashrc. No it doesn't. env runs a program in a modified environment. bash is a program where set and umask are run by bash, not env. >> There are almost always ways to work around little problems. > I think, it isn't everyone know how to solve these lttle > problems(just like me), so the best way is give the right guide in > the LFS BOOK. No, we really can't instruct for every possible combination of lack of knowledge. Having enough background in Linux to solve simple usage problems like this is really a prerequisite for LFS. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Some LFS 6.5 -> 6.6 Inconsistencies
Tim Burress wrote: > Hello, > > I've just finished upgrading from LFS 6.5 to 6.6 and there were a few > small inconsistencies I noticed along the way: > > (1) When building make, the book calls for a patch called > make-3.81-upsream_fixes-1.patch, but this is not available in the FTP > archive (or at least the mirror I used). It's there at http://www.linuxfromscratch.org/patches/lfs/6.6/make-3.81-upstream_fixes-1.patch > (2) When building perl, the book calls for a utf8 patch to avoid a > segmentation fault, but the patch is not available in the FTP archive. > By contrast, there is a libc patch in the archive that isn't in the > book. Maybe it's just a file naming problem? http://www.linuxfromscratch.org/patches/lfs/6.6/perl-5.10.1-utf8-1.patch > (3) When building e2fsprogs, the book calls for version 1.41.10 but the > FTP archive has 1.41.9. Meanwhile the e2fsprogs people recommend > 1.41.12, and the bug list is frightening enough that I just used that. OK, I'll fix up the mirror master. > And that's about it. One suggestion is that it would be REALLY, REALLY > convenient if future versions of the book had "change bars" to indicate > which parts were modified from the previous edition. There's a lot of > detailed sed hackery, etc, and it would be good to be able to see at a > glance when that's been changed (or not). That's not so easy. We create the book with Docbook xml. I don't know of a way to to do that. We do have both a change log and a "What's new" section. We are trying to do a new release every 6 months. If you look at "What's new", almost every package changes each time, but most use the instructions unchanged. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
linux fan wrote:
> # Hack to solve glibc __stack_chk_guard cross compile ( < 2.4 :)
> # this schpeel in LFS-6.6 ch5 gcc-pass1 and ch5 gcc-pass2 after unpack:
> glibc=$(ls /lib/libc-*.so)
> read j1 v[0] v[1] v[2] j2 <<< ${glibc//[-.]/ }
> if (( v[0] < 2 )) || (( v[1] < 4 )) && [ -f gcc/configure ]; then
> # kick it
> sed -i -e '/# Test for stack protector support in target C library/ {
> a\
> gcc_cv_libc_provides_ssp=yes
> } ' gcc/configure
> fi
>
> # Result: nscd with stack-smashing protector + no error in ch6 glibc
> # Notes:
> # If should probably have (( v[0] < 3 )) && (( v[1] < 4 )),
> # but could do the sed unconditionally for target LFS-6.6.
> # Don't know if needed in ch5 gcc-pass1 and/or ch5 gcc-pass2.
> # Testing on host lfs-6.2/linux-2.6.22.18,
> # "Everything is proceeding as I have foreseen it."
That's an interesting hack. Looking at gcc/configure, it seems that the
following would do the same thing.
After configure:
#ifdef TARGET_LIBC_PROVIDES_SSP
#undef TARGET_LIBC_PROVIDES_SSP
#endif
#define TARGET_LIBC_PROVIDES_SSP 1
It looks like there are three places where this is defined in gcc-4.5:
gcc/auto-host.h, prev-gcc/auto-host.h, stage1-gcc/auto-host.h
Or just an unconditional sed changing
'/* #undef TARGET_LIBC_PROVIDES_SSP */' to
'#define TARGET_LIBC_PROVIDES_SSP 1' in the above files.
I'm not 100% sure all three files there in all cases. I've been hacking
gcc in other places.
-- Bruce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: Build Runlevel
Paul Rogers wrote: > I don't recall the book saying anything about it, possibly runlevel > 3 is so typical it has never come up, but as I mentioned a week ago > or so, I think there would be an advantage to building LFS within > runlevel 2--where there's no network running, no chance of some > external attack on a vulnerable system midstream. It's easier than > "pulling the plug." Last night I discovered the perl tests really > don't like that! Some want to ping localhost, etc. I suppose it's > legitimate to expect the host to provide a protected environment, > but that newly minted LFS system really shouldn't be connected to > a network until it's "armored-up". I understand your concern, but "armored-up" is really not necessary. In order to make any type of connection to the new system, a process must be listening to a port. There are no such processes in lfs. The ping of localhost is not technically a TCP/IP process. It never goes "on the wire". After boot, the telnet, ftp, and some other clients are available, but we specifically --disable-servers in inetutils. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Some LFS 6.5 -> 6.6 Inconsistencies
Neal Murphy wrote: >>> That's not so easy. We create the book with Docbook xml. I don't know >>> of a way to to do that. We do have both a change log and a "What's new" >>> section. We are trying to do a new release every 6 months. If you look >>> at "What's new", almost every package changes each time, but most use >>> the instructions unchanged. >> Yeah, I generally work off the Changelog page when updating the scripts >> I build everything with, working up the list from whatever date I last >> built against. Works for me - Changelog open in one tab, and the index >> in a second one. > > FYI. This link might be useful: > http://www.sagehill.net/docbookxsl/Changebars.html Thanks. That may be doable, but it looks like a fairly significant effort to get all the details right. The html doesn't support change bars, but changes can be highlighted in color/underlining, etc. My thought would be to to show additions and changes, but hide deletions. I added a ticket at low priority so we don't forget about this. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
piper.guy1 wrote: > Hi, > > Started reading and doing what the book says (6.6). Didn't take too > long before I got myself into trouble. :-( > > In " Host system Requirements", the instructions explicitly wants > '/bin/sh' to be pointing to bash. Mine was pointing to dash. So I > endeavoured to change it by deleting the symlink and then create > another symlink to point to bash. However, being a naive newbie all > hell broke loose when I deleted the symlink, and everything was > misbehaving. So, before I realized what I had done I logged out > rebooted and then couldn't log back in anymore. Yes that's a problem. It's an opportunity to learn. The correct command is: ln -sfv bash /bin/sh The options are -s symbolic link -f remove existing destination file -v verbose > Sooo...before I do something else that I'm not suppose to do, I > thought I'd get advise first. My thinking is that I need to get a > Linux rescue or recovery CD, mount the file system on the hard drive, > and then add a symlink to bash. Make sense or is there an easier way? > > Any recommendations on a rescue disk? Any disk that boots to Linux. Mount the old partition and adjust the path in the ln command for the mount point. For example: # Your installed root partition is /dev/sda2 mount /dev/sda2 /mnt ln -sfv bash /mnt/bin/sh umount /mnt > One more thing. Seeing that this is a very risky thing to be advising > in LFS 6.6, can I suggest that the authour(s) add some caveats around > this instruction? LFS is not intended to teach basic Unix/Linux commands. -- bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
linux fan wrote: > I haven't read where piper.guy confirmed that bash is installed or > that if bash is not installed, that changing the link to point to bash > won't help. I don't know of a distro that doesn't install bash by default unless you are using tomsrtbt. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
Mike McCarty wrote: > # cd LFS/6.3 > # rm -rf build > > and deleted /dev from my host system! No discs, no printers, no > terminals, etc. I rebooted with a Knoppix disc, let it populate > /dev, and then mounted my hard drive, and copied (yes copied > using cp) /dev onto my hard drive. I don't think that would not be the case today since udev mounts a tempfs on top of /dev and populates it upon boot. See 6.2.2. Mounting and Populating /dev. At one time, what you did would have been needed. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
Mike McCarty wrote: > Yeah, deleting the link without changing your /etc/passwd entry > to point to a valid shell would do that. Changing the /etc/password file won't do much. The bootscripts need /bin/sh. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
Neal Murphy wrote: > I started playing with UNIX in 1986, and Linux in the > mid-nineties. And just a couple weeks ago, I overwrote a disk that contained > half of a couple striped MD filesystems. Lost nearly 10 years of pics and > history. No backups? How is this different (in effect) from a failed disk drive. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
Aleksandar Kuktin wrote: >> On Fri, 11 Jun 2010 13:16:24 -0500 >> Mike McCarty wrote: >> >> Use a red colored prompt when running with root authorization. >> >> Hope you keep learning for a long time. >> >> Mike > > See, this is a good point. :) > > I should fix this on my system. You might try this: Put this in /etc/profile and ~/.bashrc if [ -f /etc/bashrc ]; then . /etc/bashrc fi And this in /etc/bashrc NORMAL="\[\e[0m\]" RED="\[\e[1;31m\]" GREEN="\[\e[1;32m\]" if [[ $EUID == 0 ]] ; then PS1="$red\u$nor...@$hostname$red [ $NORMAL\w$RED ]# $NORMAL" else PS1="\u$gr...@$hostname$green [ $NORMAL\w$GREEN ]\$ $NORMAL" fi -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
Simon Geard wrote: > On Fri, 2010-06-11 at 13:16 -0500, Mike McCarty wrote: >> Use a red colored prompt when running with root authorization. > > Oh yes, I can't agree with that one enough. Not that you want to > accidentally run "rm -rf ~" as *any* user, I run 'rm *~' often enough. I am always careful not to get a space between the * and the ~. I suppose I ought to disable backups in vim because I never recall using one. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Errata for v6.6
On Mon, Jun 14, 2010 at 5:38 PM, James E. Lang wrote: > I am part of a Linux Users Group that has taken on LFS as a group learning > aid. > > > > The errata for Host System Requirements for v6.6 contains this line: > > Linux Kernel-2.6.22.5 (having been compiled with GCC-4.12 or greater) > > Is that GCC version a typo? Yes, that's a typo. It should be 4.1.2. I can't fix it right now as I am out of town. I'll get it fixed later this week. Thanks for letting us know. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: priority in install of packages
On Tue, Jun 15, 2010 at 4:14 AM, Mohsen Pahlevanzadeh wrote: > Dear all, > > I'm in top of Chapter 5. Constructing a Temporary System.At this chapter > some package is introduced that we must compile them.My question is: > Do they have Chapter priority in compile same listed in chapter?For > example if i compile package 5.12.expet before 5.11.tcl, is it problem? The order of the packages is carefully crafted to make sure prerequisites are built in the correct order. See the appendix for each packages dependencies. For instance, expect requires tcl. When we had a choice for the next package, it is added alphabetically, but there is really no reason to deviate from the order given in the book. You may, of course, do what pleases you, but be prepared to start over if it doesn't work. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Cannot execute grub-install
Michael Vahl wrote: > Hey y'all, > > i've recently passed all prior steps of chapter 8.4.2 (LFS 6.6). But now, > without any error messages, I'm not able to execute "grub-install > --grub-setup=/bin/true /dev/sda". grub-install can be found and sda is the > system's hdd but the system doesn't confirm the command. Nothing happens. > The only thing I can do is "Ctrl+C". Does someone else have the same problem > or can give me a hint? Do you have /dev poopulated according to section 6.2.2? What grub-install does though is copy files to /boot/grub. Is that happening? -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: AW: Cannot execute grub-install
Michael Vahl wrote: >>> Michael Vahl wrote: >>> Hey y'all, >>> >>> i've recently passed all prior steps of chapter 8.4.2 (LFS 6.6). But now, >>> without any error messages, I'm not able to execute "grub-install >>> --grub-setup=/bin/true /dev/sda". grub-install can be found and sda is > the >>> system's hdd but the system doesn't confirm the command. Nothing happens. >>> The only thing I can do is "Ctrl+C". Does someone else have the same > problem >>> or can give me a hint? > >> Do you have /dev poopulated according to section 6.2.2? > >> What grub-install does though is copy files to /boot/grub. Is that >> happening? > Thanks for your reply.. > > Unfortunately /dev is populated correctly and no grub-config files are > copied to /boot/grub: possibility one as well as possibility two I've > already considered. > > Any other suggestions? The way I would approach it would be to look at the source and see where it is hanging. grub-install is a shell script. A very simple first test would be to ensure `grub-install --help` works. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.5. GCC-4.5.0 - Pass 1
Face wrote: > is this normal > > > ../gcc-4.5.0/configure \ > --target=$LFS_TGT --prefix=/tools \ > --disable-nls --disable-shared --disable-multilib \ > --disable-decimal-float --disable-threads \ > --disable-libmudflap --disable-libssp \ > --disable-libgomp --enable-languages=c > > *** This configuration is not supported in the following subdirectories: > target-libada gnattools target-libstdc++-v3 target-libgfortran > target-libffi target-zlib target-libjava target-libobjc target-libgomp > target-libmudflap target-libssp target-boehm-gc > (Any other directories should still work fine.) It seems to be an info message. I get the same thing. All those libraries have something to do with the languages not built: c++, Objective C, FORTRAN, Java, Ada, etc. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: AW: AW: Cannot execute grub-install
Michael Vahl wrote: > After a step-by-step execution of grub-install, I determined that the script > grub-mkconfig_lib in /usr/lib/grub/ (line 35 of grub-install) doesn't have > the appropriate filemode (changed to 755). What were the permissions? They should be 644. A sourced file dies not need an executable bit. You should be running from within chroot as root at this point. > After modification an execution > of grub-mkconfig_lib caused the following stdout: /tools/bin/bash: warning: > setlocale: LC_ALL: cannot change locale (..). Anyway, I guess that the > mistake was done far earlier... Yes, the mistake was earlier. At this point, nothing should be referencing /tools. LC_ALL should be POSIX, so a change really isn't needed, but locales are installed in glibc at 6.9. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent
JAY PRAKASH SINGH wrote: > Hello , sir I am student pursuing B.tech 5th sem , I want to design design > and and implement my own operating system plz tell me how and where I should > start. Take a University level course on operating systems. Alternatively, read a book. For example: http://www.amazon.com/Operating-Systems-Design-Implementation-Second/dp/0136386776 http://www.amazon.com/IA-64-Linux-Kernel-Design-Implementation/dp/0130610143/ there are lots more... -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent
Andrew Benton wrote: > It would take hundreds of gifted people more than a decade to achieve > such a thing. Or Linus about a year. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
al...@verizon.net wrote: > Jun 20, 2010 06:21:59 AM, Andrew Benton wrote > Jun 20, 2010 08:51:09 AM, Aleksandar Kuktin wrote >> Can we see your 'config' file? > > Sure. I'll try to attach it. If you do not get it > point me to some other way to pass it along. > ... > Unfortunately (could my nouveau driver be jinxed? :): > > Jun 20, 2010 05:36:11 PM, lfs-support-boun...@linuxfromscratch.org wrote: > > Your mail to 'lfs-support' with the subject > > Re: Problem installing the "nouveau" driver > > Is being held until the list moderator can review it for approval. > > The reason it is being held: > > Message body is too big: 54361 bytes with a limit of 50 KB > > Either the message will get posted to the list, or you will receive > notification of the moderator's decision. If you would like to cancel > this posting, please visit the following URL: Use www.pastebin.org and post the resulting url. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent
Paul Rogers wrote: It would take hundreds of gifted people more than a decade to achieve such a thing. >>> Or Linus about a year. >> I'm not sure the OP could absorb that much info that quickly. > > Not to mention that Linus' first kernel wasn't what we know today. > It ran a 386, and was more an implmentation of Minix than anything > else. If there was genius in what he did, it was in throwing it > open to anybody who wanted it, and took their help in expanding it. The relationship between Linux and Minix is a common misconception. Linus did study Minix and explicitly rejected the main thrust - a microkernel. I guess he used negative knowledge; he knew what he didn't want to do. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent (Neal Murphy)
JimD. wrote: > As for Minux, another endeavor that failed to meet expectations. It depends on what you think the expectations were. My book on Minix is dated 1988 and included a 5.25 floppy with the entire source code, about 13000 lines. Tannenbaum said he wrote it for instructional reasons, not commercial, and by that standard was quite successful. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent (Neal Murphy)
JimD. wrote: > > As for Minux, another endeavor that failed to meet expectations. > >> It depends on what you think the expectations were. My book on Minix is >> dated 1988 and included a 5.25 floppy with the entire source code, about >> 13000 lines. Tannenbaum said he wrote it for instructional reasons, not >> commercial, and by that standard was quite successful. > > I agree that it depends on what your expectations are and I've heard those > goals before. > I'm sure your book is excellent, but I just don't think it would the the > ideal starting place. Ideal starting place for what? Learning how an OS works? Leaving aside the word 'ideal', the best way to learn is, IMO, with a simplified version of a complex concept that gives a general overview and enough details to start to understand the underlying complexities. Minix satisfies that goal. From that point a study of something like BSD or Linux can follow for a more in depth understanding. After all, there have been a lot of PhD Dissertations on just pieces of the system like I/O, Memory Management, Message Passing, Scheduling, Networking, etc. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent
JimD. wrote: >> Ideal starting place for what? Learning how an OS works? > The problem with Minux circa 1988 is that the code was poorly written and > buggy. > The concepts used are non-starters 20 years later. > Are you maintaining that in the past 20+ years nothing better is out there? No. Of course it has been updated. It was just and example. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: stripping down lfs
Eric Miller wrote: > hey all...haven't been here since 2002 lolz. > > I'm hoping to build a stripped down LFS to use on a livecd that will do one > thing only: present the user with a simple scripted text menu, and then > (based on the menu input) .dd an image to a usb thumb drive. > > That's it. > > boots, runs the menu, dds the chosen image to USB, then reboots (the boot > manager I'm using gives the option to boot to USB drive). > > My question: chapter six lists upwards to 60 packages...I'm sure I won't > need many if not most of them in the final product. What is the smartest > way to determine what is needed? Do I need to build complete LFS and then > strip, or can I intelligently determine what i need to build, build its > stripped down and go? If you want a really small LFS, I suggest building and stripping. For instance, gcc is not needed after the build is complete, but is is certainly needed to build everything else. Also, LFS is not designed to be the smallest possible system. There are things like busybox, uclibc, nash, etc that can reduce size considerably. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.6 Chapter 8.1 Can't "make" new kernel
Franz L. Kuhlmann wrote: > root:/sources/linux-2.6.32.8# ls > COPYING Makefile block include mm sound > CREDITS README crypto init net tools > Documentation REPORTING-BUGS drivers ipc samples usr > Kbuild arch firmware kernel scripts virt > MAINTAINERS bash fs lib security > > root:/sources/linux-2.6.32.8# make mrproper > CLEAN scripts/basic > CLEAN scripts/kconfig > CLEAN include/config > CLEAN .config include/linux/autoconf.h Try: make defconfig make to get it to build something. Then go back and do make menuconfig to customize. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.6 Chapter 8.1 Can't "make" new kernel
Neal Murphy wrote: > On Thursday 24 June 2010 16:25:45 Franz L. Kuhlmann wrote: >> ... >> ### I think the /bin/sh symlink is ok, (didn't try it in new shell yet - >> how do I do that BTW???) > > Both of the following commands display the symlink's target: > ln -s /bin/sh > readlink /bin/sh Or more conventionally: $ ls -l /bin/sh lrwxrwxrwx 1 root root 4 Aug 16 2009 /bin/sh -> bash -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: If SBU measure include test time?
littlebat wrote:
> Hi,
>
> I try to measure SBU when installing every package with command
> "time {}" and consider post my measure into
> " http://www.linuxfromscratch.org/~sbu ".
>
>
> 1, If only those packages marked "SBU testsuite included" should include
> the time spent on test into SBU measure?
>
> 2, Or, don't include the test time in SBU measure of any package?
>
> 3, Or, include every steps from the configuration, up to and including
> the first install just like the description in page
> "http://www.linuxfromscratch.org/lfs/view/stable/chapter05/binutils-pass1.html";?
>
> If I measure the SBU of "
> http://www.linuxfromscratch.org/lfs/view/stable/chapter05/tcl.html ",
> using:
>
> The first method:
> time { ./configure --prefix=/tools && make && make install; }
>
> The second:
> time { ./configure --prefix=/tools && make && make install; }
>
> The third:
> time { ./configure --prefix=/tools && make && TZ=UTC make test && make \
> install; }
>
> Which method should I use?
I don't see a difference between 1 and 2. In any case, I recommend
using times without tests in Chapert 5 and with tests in Chapter 6.
Note that the SBU site has not been updated in several years and the set
of packages on the site and in the book do not match up exactly any more.
The TZ is not needed in any case. My BLFS scripts look something like
the following, but I use jhalfs for LFS times.
-- Bruce
#!/bin/bash
source /usr/src/stats
###
# Installing program
DIR=`pwd`
PROGRAM=
LOG=$DIR/$PROGRAM.log
TITLE=$PROGRAM
TIMEFORMAT="$TIMEFMT $TITLE"
BUILDDIR=/tmp/
DEST=$BUILDDIR/install
rm -rf $BUILDDIR
mkdir $BUILDDIR
cd $BUILDDIR
before=`df -k / | grep / | sed -e "s/ \{2,\}/ /g" | cut -d' ' -f3`
tar -xf $DIR/$PROGRAM.tar.?z* || exit 1
cd $PROGRAM
{ time \
{
echo Making $TITLE
date
./configure --prefix=/usr &&
make &&
make DESTDIR=$DEST install
}
} 2>&1 | tee -a $LOG
if [ $PIPESTATUS -ne 0 ]; then exit 1; fi;
stats $LOG $DIR/$PROGRAM.tar.?z* $before
exit 0
-file stats
function stats()
{
log=$1
tarball=$2
b4=$3
free_now=`df -k / | grep / | sed -e "s/ \{2,\}/ /g" | cut -d" " -f3`
buildtime=`tail -n1 $log|cut -f1 -d" "`
sbu=`echo "scale=3; $buildtime / 132.5" | bc`
psizeK=`du -k $tarball | cut -f1`
psizeM=`echo "scale=3; $psizeK / 1024" | bc`
bsizeK=`echo "$free_now - $b4" | bc`
bsizeM=`echo "scale=3; $bsizeK / 1024" | bc`
echo "SBU=$sbu" | tee -a $log
echo "$psizeK $tarball size ($psizeM MB)"| tee -a $log
echo "$bsizeK kilobytes build size ($bsizeM MB)" | tee -a $log
(echo -n "md5sum : "; md5sum $tarball) | tee -a $log
(echo -n "sha1sum: "; sha1sum $tarball) | tee -a $log
echo "`date` $tarball" >> /usr/src/packages.log
}
TIMEFMT='%1R Elapsed Time - '
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: Issue with chroot
Saxon Landers wrote: > hi, i have just finished chapter 5 of the LFS6.6 book, and have > completed ch 6 to 6.3. I am up to chrooting into the environment. > > here's where the problem arises. the chroot command passes (is that the > right word?) the /tools/bin/env command with argument -i, however when i > run the full chroot out of the book, it says that /tools/bin/env does > not exist. I had a look in the filesystem, and indeed, it does not. Is > there something i have done wrong, how can i fix it, and, will i have to > start from scratch (fingers crossed i dont) env is installed when you build coreutils. Also make sure you followed chapter 4.2 correctly. > oh, and thankyou to all the people who helped with the glibc instalation > issue. compiling and installing in $LFS/sources fixed it. maybe the book > should specify that? It does in the -dev version of the book. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: If SBU measure include test time?
littlebat wrote: > I am learning LFS BOOK 6.6. > > Do you mean every SBU measure include test time described in the head > of the page of the package in Chapter 6 which has a testsuite and don't > include test time in Chapter 5? Yes. The SBU times are really only a guide. They are quite variable. > Have a glance at the sections in Chapter 6, I found there are some > packages which have a testsuite but without marked "testsuite included" > in the head of package section, such as: 6.25. Bison-2.4.1, > 6.33. Perl-5.10.1, 6.41. Gettext-0.17, etc... If the SBU number in > these packages don't include test time? No, we just didn't say so in those explicit sections. The times do include the tests. The book has been written by different authors over time and every detail is not specified the same. It's just not important enough to fix. If someone creates a patch, I'll apply it. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 6.7 Linux API Headers
Face wrote: > cannot extract linux-2.6.34.tar.bz2 > > > root:/sources# tar xjvf linux-2.6.34.tar.bz2 > tar (child): linux-2.6.34.tar.bz2: Cannot open: No such file or directory > tar (child): Error is not recoverable: exiting now > tar: Child returned status 2 > tar: Error is not recoverable: exiting now wget http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.34.tar.bz2 -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: If SBU measure include test time?
littlebat wrote: > On Wed, 30 Jun 2010 21:07:06 -0500 > Bruce Dubbs wrote: > > >> It's just not important enough to fix. If someone creates a patch, >> I'll apply it. > > Here is a patch, you can apply it if it is helpful. That's a script, not a patch. For an example of a patch, see http://wiki.linuxfromscratch.org/lfs/attachment/ticket/2371/program-list-update.patch -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Compilation problem for binutils
$reeHari wrote: > hi , > i've just started my lfs adventure , but unfortunately got a problem right > from the start . i'm following lfs book version 6.3 . I'm on chapter 5.3.1 - > installation of binutils . I first installed binutils and it was okey . > after that when i try recompile ld with LIB_PATH=/tools/lib , i get error . > help me on this guys .. i use ubuntu 10.04 (if dat can be of ny help ) . > but in version 6.6 of the book , the ld is not recompiled as i see it . > isn't it necessary to change the library search path ? > I tried the following and i got error > > make -C ld clean > make -C ld LIB_PATH=/tools/lib > > this is the error i get : > > make: Entering directory > `/home/jacsparrow/LINUX_FROM_SCRATCH/sources/binutils-build/ld' There are several issues here. First, virtually nobody remembers details from lfs 6.3. That was released in 2007. The current version is 6.6. Second, you are building on /home/jacsparrow/LINUX_FROM_SCRATCH/sources/, but that will cause you problems when you get to Chapter 6. You will not be able to access the sources when you chroot to /mnt/lfs/. Put the sources in /mnt/lfs/sources and build from there. Third, make sure you've got all the packages you need as specified in the Host System Requirements. There are several changes needed in most Ubuntu systems. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: If SBU measure include test time?
littlebat wrote: > On Thu, 01 Jul 2010 10:41:38 -0500 > Bruce Dubbs wrote: >> That's a script, not a patch. >> >> For an example of a patch, see >> http://wiki.linuxfromscratch.org/lfs/attachment/ticket/2371/program-list-update.patch > > Here is a patch, you can apply it if it is helpful. > > Ticket #2702 (new enhancement) clarifying the approximate build SBU > number and required disk space of the package > http://wiki.linuxfromscratch.org/lfs/ticket/2702 > > patch file: > http://wiki.linuxfromscratch.org/lfs/attachment/ticket/2702/clarify-sbu-disk_space.patch You don't need to repeat ticket info here. I see all tickets as they are posted. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
Neal Murphy wrote: > On Friday 02 July 2010 15:11:31 Face wrote: >> well, thank you all, i will start over and see what will happen. >> I am using a shell script to do the book if someone could take a look >> at it, that would be nice . -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
Face wrote: > well, thank you all, i will start over and see what will happen. > I am using a shell script to do the book if someone could take a look > at it, that would be nice . Ignore the previous post I accidentally sent. For any complex task like LFS, you should be able to successfully do the operations manually before trying to script it. If you insist on scripting, I highly recommend a separate script for each page rather than a monolithic 1000+ line script. As a comparison, jhalfs scripts (all 30 for Chapter 5) total 410 lines. The Makefile, which covers the entire book is 1700 lines... -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
Face wrote: > it has been over 2 months since my first try i think. going through > the book manually over and over, I had many mistypes which force me to > start over. Therefore, I put them in a script so i can avoid mistypes. > Yes i do agree the script is way too long , however, alot of the lines > is about tell me where i am At and what is going on in a way i can > understand and each chapter is in a separate function. if you cut the > unnecessary echo's lines you will end up with a script that has less > than 200 lines. Your distro, your rules. You don't have to justify anything to me. However, you did ask for advice. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.6 Chapter 2, Preparing a New Partition
Ken Moffat wrote: > On 3 July 2010 14:28, linux fan wrote: >> If the 50GB host Operating System(Debian 5.0.5) is needed to run (yes, >> I think), then you need to use a separate partition to contain the >> destination LFS system. I think 12 GB is a good size. 5GB is a little >> too small if many things will be installed (a typical desktop system >> with lots of goodies). > > I shudder to think about what you are using to fill that up ;) I'm > still happy with 3GB for '/', with separate /sources and /home. > Your system, your rules - just make sure you remember to back > up. On a machine I've been using for several years: FilesystemSize Used Avail Use% Mounted on /dev/sda5 8.1G 6.8G 970M 88% / /dev/sda13 17G 6.1G 9.4G 40% /opt On my new non-lfs system: /share 97T 5.0T 92T 6% /share /work 194T 167T 27T 87% /work /scratch 806T 685T 121T 85% /scratch /corral 688T 235T 419T 36% /corral BTW: $ free total used free Mem: 32913636 32844280 69356 ... processor : 15 vendor_id : AuthenticAMD cpu family : 16 model : 2 model name : Quad-Core AMD Opteron(tm) Processor 8354 stepping: 2 cpu MHz : 2200.000 :) -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Question about type of partition to use
James Butler wrote: > Can the dedicated partition that I use for my lfs be a logical volume created > from a dedicated partition? > This is what my setup looks like: > > /dev/sda1/boot > > Volume Groups > >vg_deltaflyer > >Physical View > >/dev/sda2 > >Logical View > lv_root > lv_swap > >vg_deltaflyer00 > >Physical View > >/dev/sda3 > >Logical View > lv_lfs > > > Partitions 1 and 2 hold the host system and partition 3 is for the LFS system. > Will this work? You can probably build, but booting a standard LFS system will be a problem. Linux doesn't understand logical volumes without an initrd and LFS does not address that. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Boot kernal Panic :(
Saxon Landers wrote: > "" > I'm making a *swag* on this. I'll bet your 'sdb' is called 'sdb1' or has > a uuid given to it by the Ubuntu kernel. I'm not very sure I can help, but the uuid is created by mkfs, when the partition is formatted under whatever distro you are using. > List of all partitions: > 0800 321571224 sda driver: sd > 0801 1228800 sda1 > 0802 211340376 sda2 > 0b00 1048575 sr0 driver: sr > No filesystem could mount root, tried: ext3 vfat msdos iso9660 > Kernal Panic - not syncing VFS: unable to mount root fs on > unknown-block(8,0) I don't see what you used in grub to start the kernel, but it is obviously finding the kernel. There are two potential problems. Either you don't have the ext2/3 filesystem built into the kernel or the root= parameter on the kernel line does not specify what it needs. You should have in the kernel config CONFIG_EXT3_FS=y That is probably OK because the kernel tried it. (Note the only difference between ext2 and ext3 is the journal, so ext3 is a superset of ext2.) Looking at the above, I'd say that you should specify root=/dev/sda1 or root=/dev/sda2. You may also have a kernel config problem with not configuring with the proper usb devices -- don't set anything to modules. I'd also check a working disto's logs to see what USB parameters are being initialized and make sure those are set in the kernel config. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
