0.48 - Can't mount /proc
Erik,
No, doesn't help.
bash# mount proc /proc -t proc
mount: Mounting proc on /proc failed: Unknown error 716878944
Maybe people in mips-linux know something about this?
Quoted 20 lines · Erik Andersen
Quoted 9 lines · Michael Shmulevich
Hello,
This is my first mail to this list, but I have already seen similar question being unanswered.
I am developing an embedded system based on MIPS board. I am cross-compiling the BusyBox 0.48 on x86 with gcc 2.90.29 (egcs 1.0.3a).While trying to mount /proc either atomatically in script or by hand, the command fails with following message:
bash# mount /proc -t proc
mount: Mounting none on /proc failed: Unknown error 716878944mount proc /proc -t proc
-Erik
--
Erik B. Andersen email: [email protected]
--This message was written using 73% post-consumer electrons--
Quoted 10 lines · Michael Shmulevich
Erik,
No, doesn't help.
bash# mount proc /proc -t proc
mount: Mounting proc on /proc failed: Unknown error 716878944
Maybe people in mips-linux know something about this?
Yes, this does sound like a kernel specific problem then, since this works on (at least) arm, ppc, sh, and x86. I havn't tried it on anything else.
-Erik
--
Erik B. Andersen email: [email protected]
--This message was written using 73% post-consumer electrons--
Quoted 7 lines · Erik Andersen
Quoted 6 lines · Michael Shmulevich
Erik,
No, doesn't help.
bash# mount proc /proc -t proc
Quoted 8 lines
Quoted 4 lines
mount: Mounting proc on /proc failed: Unknown error 716878944
Maybe people in mips-linux know something about this?
Yes, this does sound like a kernel specific problem then, since this works on (at least) arm, ppc, sh, and x86. I havn't tried it on anything else.
Henning Heinold has done a lot with busybox on mips (debian installer) Might this be due to kernel 2.4 and /proc/mounts includeing a / in front of proc
(flo@ping)~# cat /proc/mounts /dev/root / ext2 rw 0 0 proc /proc proc rw 0 0 devpts /dev/pts devpts rw 0 0 /dev/sda7 /usr ext2 rw 0 0 /dev/sda8 /var ext2 rw 0 0 /dev/sda9 /tmp ext2 rw 0 0 /dev/sda10 /home ext2 rw 0 0 /dev/hda5 /var/tmp ext3 rw 0 0 usbdevfs /proc/bus/usb usbdevfs rw 0 0 automount(pid19472) /mnt autofs rw 0 0 (flo@ping)~# uname -a Linux ping.mediaways.net 2.2.18ext3 #1 Thu Dec 14 18:24:45 CET 2000 i686 unknown
Or
flo@resume:~$ cat /proc/mounts /dev/root / ext2 rw 0 0 /proc /proc proc rw 0 0 /dev/sdb1 /home2 ext2 rw 0 0 /dev/sdc1 /home3 ext2 rw 0 0 /dev/sdd1 /ftp.rfc822.org ext2 rw 0 0 /dev/sde1 /home4 ext2 rw 0 0 shmfs /var/shm shm rw 0 0 devpts /home4/dev/pts devpts rw 0 0 proc /home4/proc proc rw 0 0 flo@resume:~$ uname -a Linux resume.rfc822.org 2.4.0-test6 #2 Sun Aug 27 12:37:51 GMT 2000 mips unknown
Henning has fought with this IIRC.
Flo
Signature · 2 lines
Florian Lohoff [email protected] +49-5201-669912 Why is it called "common sense" when nobody seems to have any?
Here's a kernel patch. The __access_ok macro looks one byte too far and fails. Since copy_mount_options() isn't sure how long the string arguments are, it just copies to the end of the page. Since this is on busybox's stack, the copy wants to go all the way to 0x7FFFFFF and hits this corner case.
Quinn Jensen
[email protected]
include/asm-mips/uaccess.h Wed Jan 10 16:20:35 2001 renamed from include/asm-mips/uaccess.h Wed Oct 4 19:19:02 2000 +1 -1
@@ -46,7 +46,7 @@ * - OR we are in kernel mode. */ #define __access_ok(addr,size,mask) \ - (((__signed__ long)((mask)&(addr | size | (addr+size)))) >= 0) + (((__signed__ long)((mask)&(addr | size | (addr+size-1)))) >= 0) #define __access_mask ((long)(get_fs().seg)) #define access_ok(type,addr,size) \
Quoted 3 lines · Michael Shmulevich
bash# mount proc /proc -t proc
mount: Mounting proc on /proc failed: Unknown error 716878944
716878944 is 0x2abab460, a typical address of a shared library or malloced memory allocated via mmap(2).
Ralf
Quoted 6 lines · Quinn Jensen
Here's a kernel patch. The
__access_okmacro looks one byte too far and fails. Sincecopy_mount_options()isn't sure how long the string arguments are, it just copies to the end of the page. Since this is on busybox's stack, the copy wants to go all the way to0x7FFFFFFand hits this corner case.
I don't like this solution as it inflates the kernel noticably. Actually even the bug itself hasn't been one; this off by one mistake was deliberatly accepted in the - obviously wrong - assumption that nobody would ever try to use the last byte of userspace. See also the Alpha variant of the code; looks like they suffer from the same problem.
My solution will be to truncate userspace by by at least 4kb. I've choosen to even truncate it by 32kb; this will also make the layout of the address space for 32-bit processes on 64-bit kernels and 32-bit kernel identical again.
Ralf