PDA

View Full Version : probe partition for geom failed


tastefulasever
5th October 2004, 01:50 AM
After choosing type of keyboard layout during fedora 3 test installation, i recieve the following errors/bugs.

"assertion (heads<256) at disk_dos.c: 472 in function probe_partition_for_geom() failed."

"Assertion (sectors <= 63) at disk_dos.c:476 in function probe_partition_for_geom() failed."

"Assertion (heads>0) at disk_dos.c:476 in function probe_partition_for_geom() failed."

I imagine it has something to do with the current format of my harddrive, or the fact that i have win2000 and 98 installed, but im completely new to linux.

i have one 40 gig harddrive, 10gigs for win2000, 20gigs for storage, 5 gigs for win98, and the rest unpartitioned.

Any help would be appreciated.


Open Source is the coolest idea ever..

Shadow Skill
6th October 2004, 06:36 PM
I have the same sort of problem with my laptop, I am not sure what I did to break the drive geometry or how to fix it though. I think I got this problem after using multiple programs to format the HDD of my mobile and frequent reimaging on my part.

LordMorgul
6th October 2004, 07:08 PM
It would be very handy if you could gather some information about your disk partition table. This may be alot to write down, and you may not have a way to save the output, but here is what you could try.

As soon as the installer starts, before the Assertion errors show up, push CTRL-ALT-F2 and you will be shown a command prompt logged in. Here you want to run the fdisk command, and ask it to print out the information about your disk partition table. Knowing everything this command prints out about your setup can help determine what is wrong, or give you steps to work around the problem.

# /sbin/fdisk -l /dev/hda

The # is not part of the command. The flag -l is the letter L lowercase. This assumes your harddrive is operating on hda (the first ide controller). If it is not, then you might need to use /dev/hdb, or perhaps you do not have an ide drive and /dev/sd0 (for sata / scsi) is what you need.

Be careful with this program, if it loads and shows you a menu rather than just printing the partition table and exiting, push q to quit.

I saw this assertion error myself, but my disk layout is ridiculous. The disk was partitioned with WindowsXP disk manager, and that is part of the problem.

rickrich
16th October 2004, 05:35 AM
Yeah, I got the "assertion (heads<256) at disk_dos.c: 472 in function probe_partition_for_geom() failed."
message over and over trying to install FC3T2 on a new Compaq R3000Z. I kept ignoring the message. Then after package selection was complete, I got a fatal traceback from python and then the X display went into "TV on the fritz" mode.

Its really frustrating when basic things like this, which worked well for years in the RHx series, get broken in Fedora.

rickrich
16th October 2004, 05:51 AM
It would be very handy if you could gather some information about your disk partition table. This may be alot to write down, and you may not have a way to save the output, but here is what you could try.

As soon as the installer starts, before the Assertion errors show up, push CTRL-ALT-F2 and you will be shown a command prompt logged in. Here you want to run the fdisk command, and ask it to print out the information about your disk partition table. Knowing everything this command prints out about your setup can help determine what is wrong, or give you steps to work around the problem.

# /sbin/fdisk -l /dev/hda


Unfortunately, on the Compaq R3000Z, switching to one of the text mode VT's from the graphical installer just results in a rolling screen out of sync. So I did it from a text mode install. This is what I got:


Disk /dev/hda: 40.0 GB, 40007761920 bytes
255 heads, 63 sectors/track, 4864 cylinders
Units = cylinders of 16065 * 512 = 8225280

Device Start End Blocks Id System
/dev/hda1 1 2549 20474811 7 HPFS/NTFS
/dev/hda2 2550 2562 104422+ 83 Linux
/dev/hda3 2563 4864 18490815 8e Linux LVM

LordMorgul
16th October 2004, 11:25 AM
This bug should be fixed in newer versions of Parted, see these bugs:
https://bugzilla.redhat.com/beta1/buglist.cgi (probe_partition_for_geom) (https://bugzilla.redhat.com/beta1/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&product=Fedora+Core&version=devel&version=fc3test1&version=fc3test2&version=fc3test3&component_text=&query_format=advanced&bug_status=NEW&bug_status=VERIFIED&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=RESOLVED&bug_status=CLOSED&bug_status=NEEDINFO&bug_status=MODIFIED&long_desc_type=substring&long_desc=probe_partition_for_geom&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&fixed_in_type=allwordssubstr&fixed_in=&keywords_type=allwords&keywords=&emailtype1=exact&email1=&emailtype2=exact&email2=&bugidtype=include&bug_id=&votes=&changedin=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Last+Changed&remtype=asdefault&field0-0-0=noop&type0-0-0=noop&value0-0-0=)

btw: I remember this issue in both RH 8 and RH9 with abnormal partition tables (meaning, those created by windows).. it happens in tests. :-)

If you're not trying FC3test3 yet, see if you get the same warnings with the new version, the problem was resolved for me. If not, report to these bugs that the issue did not go away in test3.

peter@lorenzen
25th October 2004, 03:05 PM
This bug should be fixed in newer versions of Parted, see these bugs:
https://bugzilla.redhat.com/beta1/buglist.cgi (probe_partition_for_geom) (https://bugzilla.redhat.com/beta1/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&product=Fedora+Core&version=devel&version=fc3test1&version=fc3test2&version=fc3test3&component_text=&query_format=advanced&bug_status=NEW&bug_status=VERIFIED&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=RESOLVED&bug_status=CLOSED&bug_status=NEEDINFO&bug_status=MODIFIED&long_desc_type=substring&long_desc=probe_partition_for_geom&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&fixed_in_type=allwordssubstr&fixed_in=&keywords_type=allwords&keywords=&emailtype1=exact&email1=&emailtype2=exact&email2=&bugidtype=include&bug_id=&votes=&changedin=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Last+Changed&remtype=asdefault&field0-0-0=noop&type0-0-0=noop&value0-0-0=)

btw: I remember this issue in both RH 8 and RH9 with abnormal partition tables (meaning, those created by windows).. it happens in tests. :-)

If you're not trying FC3test3 yet, see if you get the same warnings with the new version, the problem was resolved for me. If not, report to these bugs that the issue did not go away in test3.

Hi

I have the same problem trying to install FC3-test on a system that previously was running FC1.

The problem is: parted hits a bug when operating on one of the harddisk.
I tried to install parted-1.6.15-5.i386.rpm but the problem is still there,
fdisk on the otherhand works fine on that disk
if there is any data I can give you please let me know

/Peter@Lorenzen.us

peter@lorenzen
26th October 2004, 04:58 PM
Hi

I have the same problem trying to install FC3-test on a system that previously was running FC1.

The problem is: parted hits a bug when operating on one of the harddisk.
I tried to install parted-1.6.15-5.i386.rpm but the problem is still there,
fdisk on the otherhand works fine on that disk
if there is any data I can give you please let me know

/Peter@Lorenzen.us
Hi again

I have made this fix to parted-1.6.15-5, which maked parted fallback and use
fdisk method of establishing what C/H/S should be, there is a few ekstra
printf's, so that some-one can figure out what actually was wrong.

I have included 3 parts, the patch, output from fdisk, output from the patched parted

I hope it will be of help to somebody

/Peter@Lorenzen.us

[root@fire ~]# diff -Naur parted-1.6.15*/libparted/disk_dos.c
--- parted-1.6.15/libparted/disk_dos.c 2004-09-19 01:37:18.000000000 +0100
+++ parted-1.6.15-modified/libparted/disk_dos.c 2004-10-26 17:37:33.589363808 +0100
@@ -472,17 +472,54 @@
return 0;

heads = (A_ * h - a_ * H) / (a_ * C - A_ * c);
+#define ALA_FDISK 1
+#ifdef ALA_FDISK
+/*
+ * I don't understand the meaning of the above statement
+ * But since parted is showstopper in installing FC3-test
+ * I have made this FIX - taht is we do behave like fdisk
+ * peter@lorenzen.us 26/10-2004
+ */
+ printf("c = %d, ", c); printf("h = %d, ", h); printf("s = %d, ", s);
+ printf("a = %d, ", a); printf("a_ = %d\n", a_);
+
+ printf("C = %d, ", C); printf("H = %d, ", H); printf("S = %d, ", S);
+ printf("A = %d, ", A); printf("A_ = %d\n", A_);
+
+ if ((heads <= 0) || (heads >= 255)) {
+ /* fdisk:get_partition_table_geometry() h = p->end_head + 1 */
+ heads = H + 1;
+ /* fdisk:get_partition_table_geometry() s = p->end_end_sector
+ note that chs_get_sector() for some reason subtracts one */
+ sectors = S + 1;
+ }
+#endif
PED_ASSERT (heads > 0, return 0);
PED_ASSERT (heads < 256, return 0);

sectors = A_ / (C * heads + H);
+#ifdef ALA_FDISK
+ if ((sectors <= 0) || (sectors > 63)) {
+ /* fdisk:get_partition_table_geometry() h = p->end_head + 1 */
+ heads = H + 1;
+ /* fdisk:get_partition_table_geometry() s = p->end_end_sector
+ note that chs_get_sector() for some reason subtracts one */
+ sectors = S + 1;
+ }
+#endif
PED_ASSERT (sectors > 0, return 0);
PED_ASSERT (sectors <= 63, return 0);

cylinders = part->disk->dev->length / (heads * sectors);

+#ifdef ALA_FDISK
+ printf("cylinders = %d\n", cylinders);
+ printf("heads = %d\n", heads);
+ printf("sectors = %d\n", sectors);
+#else
PED_ASSERT ((c * heads + h) * sectors + s == a, return 0);
PED_ASSERT ((C * heads + H) * sectors + S == A, return 0);
+#endif

bios_geom->cylinders = cylinders;
bios_geom->heads = heads;
[root@fire ~]#
/sbin/fdisk -lu /dev/hde

Disk /dev/hde: 46.1 GB, 46115758080 bytes
255 heads, 63 sectors/track, 5606 cylinders, total 90069840 sectors
Units = sectors of 1 * 512 = 512 bytes

Device Boot Start End Blocks Id System
/dev/hde1 63 16064 8001 83 Linux
/dev/hde2 * 16065 90060389 45022162+ 83 Linux
[root@fire ~]# ./parted-1.6.15-modified/parted/parted /dev/hde print
c = 1, h = 0, s = 0, a = 16065, a_ = 16065
C = 485, H = 254, S = 62, A = 90060389, A_ = 90060327
cylinders = 5606
heads = 255
sectors = 63
c = 1, h = 0, s = 0, a = 16065, a_ = 16065
C = 485, H = 254, S = 62, A = 90060389, A_ = 90060327
cylinders = 5606
heads = 255
sectors = 63
Error: Unable to satisfy all constraints on the partition.
Error: Can't have the end before the start!
Disk geometry for /dev/hde: 0.000-43979.414 megabytes
Disk label type: msdos
Minor Start End Type Filesystem Flags
1 0.031 7.844 primary ext2
2 7.844 43974.799 primary ext3 boot
Information: Don't forget to update /etc/fstab, if necessary.

[root@fire ~]#

vnpenguin
9th November 2004, 07:58 AM
I have the same error with FC3 (not test). In the same harddisk I installed without any problem with FC1, FC2, Slackware 10.0, SUSE 9.1,...

So I cant install FC3 :(

Any solution ?

Menden Photos on Instagram - Ebebiyin - Al Battaliyah Photos