Kernel Compile v2.4.20-SuSE-100
I'm not sure that I've got the kernel version completely right, it's the latest available version from YOU. I compile the kernel and put it in /boot/ as bzImage.Olipro, I create a boot entry in grub but I get an error every time it starts to boot, first error is that the reiserfs module is not found and thus results in a kernel panic that it could not mount root. (I kept it all separate so I wouldn't have to go through a long-winded recovery process in the event something went wrong) just to ensure I'm not doing something wrong, here's the series I follow to compile: bash#: cd /usr/src/linux bash#: make clean # not mandatory I know bash#: make mrproper bash#: make clean # not mandatory I know bash#: make menuconfig # (using SuSE config: CPU changed to Pentium-4) bash#: make dep bash#: vi Makefile # I modify the EXTRAVERSION parameter to -01.Olipro bash#: make bzImage bash#: make modules bash#: make modules_install That's it then I move my bzImage to /boot/ and add an entry to GRUB. (I sense those bash commands are useless but maybe it'll help someone who doesn't know how to compile their kernel) --------------------------------- Olipro - GTA Global <Olipro@GTAGlobal.net>
--=.c:Y8C(tBMwOhnm Content-Type: text/plain; charset=US-ASCII Content-Disposition: inline Content-Transfer-Encoding: 7bit On 03 Sep 2003 21:57:11 +0100 Olipro <Olipro@GTAGlobal.net> wrote:
first error is that the reiserfs module is not found and thus results in a kernel panic that it could not mount root.
You compiled reiserfs as a module so you need to create an initrd with mkinitrd or recompile reiserfs into the kernel. Charles -- "The world is beating a path to our door" -- Bruce Perens, (Open Sources, 1999 O'Reilly and Associates) --=.c:Y8C(tBMwOhnm Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (GNU/Linux) iD8DBQE/Vldz3epPyyKbwPYRAuxrAJ4uyU3VrJFDomyT3ThFyygTsCzCzwCgrmUB 5HOQ4gLRG44NRRBKD1z+saA= =sgEe -----END PGP SIGNATURE----- --=.c:Y8C(tBMwOhnm--
The 03.09.03 at 17:04, Charles Philip Chan wrote:
[ Malformed message. Use "H" to see raw text. ]
I can not see your email; not only this one, but several of yours. Pine shows the above line, and no text (except in raw mode). Mozilla and Netscape shows nothing (unless I look at source), and balsa says "multipart message has no boundary parameter" and shows no message text (except under show raw text). If I look at the raw text, I see strange headers with escapes: X-Face: G;Z,`sm>)4t4LB/GUrgH$W`!AmfHMj,LG)Z}X0ax@s9:0>0)B&@vcm{v-le)wng)?|o]D<V6&ay<F=H{M5?$T%p!dPdJeF, au\E@TA"v22K!Zl\\mzpU4]6$ZnAI3_L)h;fpd}mn2py/7gv^|*85-D_f:07cT>\Z}0:6X Here, it should say what is the boundary mark, but there is no "boundary" keyword. Content-Type: multipart/signed; protocol="application/pgp-signature"; micalg="pgp-sha1";: Y8C(tBMwOhnm" Then, later, when the boundary appears, it doesn't match the above: --=.c:Y8C(tBMwOhnm See? The 'c' doesn't appears previously. That version of Sylpheed must be broken, I think. -- Cheers, Carlos Robinson
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thu, 4 Sep 2003 00:51:11 +0200 (CEST) "Carlos E. R." <robin1.listas@tiscali.es> wrote:
See? The 'c' doesn't appears previously. That version of Sylpheed must be broken, I think.
I noticed that too, it seems like once it a while the "boundary" keyword is not there. Unfortunately sourceforge is having problems with their CVS server again- I am waiting for an update which should solve the problem. I didn't know that people are having problems reading them since Sylpheed-claws does show the text. Sorry for the inconvenience, I have switched to clear signing for now. Charles - -- "However, complexity is not always the enemy." -- Larry Wall (Open Sources, 1999 O'Reilly and Associates) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (GNU/Linux) iD8DBQE/Vn5s3epPyyKbwPYRAohJAJ9Jnb5qbJ8tltv2CsiTygQAr9W2SgCeO4ld fxtpzZ8s4WnMHGuQLafR2m0= =Vk/j -----END PGP SIGNATURE-----
The 03.09.03 at 19:51, Charles Philip Chan wrote:
I noticed that too, it seems like once it a while the "boundary" keyword is not there. Unfortunately sourceforge is having problems with their CVS server again- I am waiting for an update which should solve the problem. I didn't know that people are having problems reading them since Sylpheed-claws does show the text. Sorry for the inconvenience, I have switched to clear signing for now.
It doesn't affect all your emails, only some. Some of the good ones are signed as well, so I don't really know how/when exactly it happens. I can read them in raw mode, no big problem: but I thought I should let you know there is a problem somewhere. -- Cheers, Carlos Robinson
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thu, 4 Sep 2003 02:04:42 +0200 (CEST) "Carlos E. R." <robin1.listas@tiscali.es> wrote:
It doesn't affect all your emails, only some.
I know, that is why I said once in a while.
so I don't really know how/when exactly it happens.
I can't find the pattern either. :-( Charles - -- "However, complexity is not always the enemy." -- Larry Wall (Open Sources, 1999 O'Reilly and Associates) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (GNU/Linux) iD8DBQE/VoUk3epPyyKbwPYRAnsEAKDXuyTpgXJlM9C+YJ0yQhoSzfln7QCgu7W0 QgAIYu6ivf6B/k6GgwVVDhE= =eSVH -----END PGP SIGNATURE-----
On Wed, 3 Sep 2003 20:19:49 -0400 Charles Philip Chan <cpchan@sympatico.ca> wrote:
I can't find the pattern either. :-(
OK, I found out the problem is not Sylpheed-claws. Now I am leaning towards sympatico doing some content filtering and mistook the string for a virus. Charles -- "Are [Linux users] lemmings collectively jumping off of the cliff of reliable, well-engineered commercial software?" (By Matt Welsh)
The 03.09.03 at 21:48, Charles Philip Chan wrote:
OK, I found out the problem is not Sylpheed-claws. Now I am leaning towards sympatico doing some content filtering and mistook the string for a virus.
Could be... your headers seems strange to me, with strange chars. It could confuse a filter, perhaps. At first I thought that more people from that server had the same problem, but I'm unsure. You see, I have a cousin using an account there, so I tend to notice emails from that server, but not the user name :-) Interesting... They do outgoing email filtering? And also incoming, I suppose. That's nice of them - or perhaps they charge extra? -- Cheers, Carlos Robinson
On Thu, 4 Sep 2003 14:54:33 +0200 (CEST) "Carlos E. R." <robin1.listas@tiscali.es> wrote:
Interesting... They do outgoing email filtering? And also incoming, I suppose. That's nice of them - or perhaps they charge extra?
I don't really know for sure. I am guessing here since I figured out that it is not a Sylpheed problem and I haven't changed my setup. Charles -- Linux: Because a PC is a terrible thing to waste. (By komarimf@craft.camp.clarkson.edu, Mark Komarinski)
On Thu, 4 Sep 2003 10:04:25 -0400 Charles Philip Chan <cpchan@sympatico.ca> wrote:
I don't really know for sure. I am guessing here since I figured out that it is not a Sylpheed problem and I haven't changed my setup.
I have solved my problem. Now it seems like some smtp server chokes on : in the tag. Sylpheed have now been changed not to allow that. Charles -- "...and scantily clad females, of course. Who cares if it's below zero outside" (By Linus Torvalds)
The 03.09.04 at 10:37, Charles Philip Chan wrote:
I have solved my problem. Now it seems like some smtp server chokes on : in the tag. Sylpheed have now been changed not to allow that.
Why does it use ":"? I would think that is a reserved char in the headers, it could break many filters. -- Cheers, Carlos Robinson
On Fri, 5 Sep 2003 00:16:29 +0200 (CEST) "Carlos E. R." <robin1.listas@tiscali.es> wrote:
Why does it use ":"? I would think that is a reserved char in the headers, it could break many filters.
The string is generated randomly, it seems. : is one of the allowed characters according to RFC 2015 which governs multipart mime. Charles -- A Linux machine! because a 486 is a terrible thing to waste! (By jjs@wintermute.ucr.edu, Joe Sloan)
The 03.09.04 at 18:48, Charles Philip Chan wrote:
Why does it use ":"? I would think that is a reserved char in the headers, it could break many filters.
The string is generated randomly, it seems. : is one of the allowed characters according to RFC 2015 which governs multipart mime.
I'm no expert in that, so I'll take your word on it :-) . But I think that char inside a header can cause problems, because it is used as a separator. -- Cheers, Carlos Robinson
--cp_:eqDxG,g=.qK5 Content-Type: text/plain; charset=US-ASCII Content-Disposition: inline Content-Transfer-Encoding: 7bit Correction it should be multipart PGP MIME. Charles -- "All language designers are arrogant. Goes with the territory..." (By Larry Wall) --cp_:eqDxG,g=.qK5 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (GNU/Linux) iD8DBQE/V8IF3epPyyKbwPYRAknRAKCUGTnrB1ab4xUUen62F8hlJsmcVQCfV6WQ UKubSW3vpnrdKn68uZSJkAA= =NNxi -----END PGP SIGNATURE----- --cp_:eqDxG,g=.qK5--
The 03.09.04 at 18:51, Charles Philip Chan wrote:
[ Malformed message. Use "H" to see raw text. ]
Another one... it is not solved yet.
Content-Type: multipart/signed; protocol="application/pgp-signature"; micalg="pgp-sha1";: eqDxG,g=.qK5"
Correction it should be multipart PGP MIME.
But the problem is the header. That one above is incorrect, I think: the "boundary" should have been there, and it isn't. This is another one from you, that shows well:
Content-Type: multipart/signed; protocol="application/pgp-signature"; micalg="pgp-sha1"; boundary="=.QBKod..B+OrXBD"
______________________^^^^^^^ -- Cheers, Carlos Robinson
On Fri, 5 Sep 2003 01:12:00 +0200 (CEST) "Carlos E. R." <robin1.listas@tiscali.es> wrote:
Another one... it is not solved yet.
I am in the process of compiling the program. Sourgeforge CVS haven't been updated in days. One of the developers was kind enough to send me a patch though. Charles -- "If you want to travel around the world and be invited to speak at a lot of different places, just write a Unix operating system." (By Linus Torvalds)
On Thu, 4 Sep 2003 10:04:25 -0400 Charles Philip Chan <cpchan@sympatico.ca> wrote:
Interesting... They do outgoing email filtering?
I was right. I found out from another sympatico user on the Sylspheed list that indeed they are filtering mail and is messing his emails too. Apparently, sympatico, ironically, is using brigtmail. I think that company should be renamed dimmail. Charles -- "All language designers are arrogant. Goes with the territory..." (By Larry Wall)
The 03.09.03 at 21:57, Olipro wrote:
just to ensure I'm not doing something wrong, here's the series I follow to compile:
bash#: cd /usr/src/linux bash#: make clean # not mandatory I know bash#: make mrproper
mrproper is not needed normally: it clears too many things. It is used sometimes after expanding sources.
bash#: make clean # not mandatory I know
Redundant, after mrproper.
bash#: make menuconfig # (using SuSE config: CPU changed to Pentium-4)
Ok, no. Before that, use "make cloneconfig", because it is a SuSE kernel. This will clone the configuration of the running kernel - which will be probably the binary version of same, or very similar - ensuring your compiled version is equivalent and minimizes your chances of error. Then, use menuconfig and change the cpu, as you did.
bash#: make dep bash#: vi Makefile # I modify the EXTRAVERSION parameter to -01.Olipro
Ok, I do that before all the above (except the mrproper). I'm not sure the order matters, but just in case.
bash#: make bzImage bash#: make modules
You can save some time by typing "make dep bzImage modules" on a single line: make is clear enough, and as it is slow i can go away for a coffe without the need to enter the next command ;-)
bash#: make modules_install
That's it then I move my bzImage to /boot/ and add an entry to GRUB.
I usually use make install instead. It also saves vmlinuz to vmlinuz.old. I also keep vmlinuz.shipped, with the corresponding grub entry. -- Cheers, Carlos Robinson
Olipro wrote:
I compile the kernel and put it in /boot/ as bzImage.Olipro, I create a boot entry in grub but I get an error every time it starts to boot, first error is that the reiserfs module is not found and thus results in a kernel panic that it could not mount root. (I kept it all
Did you also create a new initial ramdisk image for your new kernel? If you have kept the default SuSE kernel configuration (apart from changing CPU type) reiserfs is built as a module and therefore initrd is required. -- Martti Laaksonen
participants (4)
-
Carlos E. R. -
Charles Philip Chan -
Martti Laaksonen -
Olipro