Hi Ernest,
Ernest jw ter Kuile wrote:
Oops bug
Version 448 discovers a partitionless card while there most certainly is a
partition.
I should know, I just put it there, and mounted it under linux (its /dev/sda5).
Your detection routine needs tweaking.
What do you check for to see if it has a partition ? just before your error
message I see the normal Linux message displaying (copied by hand):
Synthesizing the initial hotplug events
hda: SAMSUNG CF/ATA, CFA DISK drive
ide0 at 0xc4820000-0xc4820007,0xc482000e on irq 92
hda: max request size 128KiB
hda: 2041200 sectors (1045 MB) w/0KiB Cache, CHS=2025/16/63
hda: hda1 <hda5 hda6 hda7 >
ide-cs: hda: Vcc = 3.3, Vpp = 0.0
Waiting for /dev ...
ending with your :
ERROR: PARTITIONLESS CARD etc...
yes, my mistake, fixed in r449+ and about to rebuild+upload
btw: I've almost completed writing a script to convert a partionless
card into a partitioned one (the whole issue annoys me TBH;). I'll put
it through some rigorous testing and see if I can prove that it
works. Aiming to include it in 0.6.4 if I can get it rock solid.
Ernest ter Kuile.
ps: Would you mind not screaming like that ?
?
On Tuesday 26 September 2006 02:00, Paul Sokolovsky wrote:
Greetings,
Friday, September 22, 2006, 10:33:01 PM, you wrote:
On Thursday 21 September 2006 01:54, you wrote:
Hmmm ... hadn't thought of that possibility.
Horrible way of doing it IMHO. PPC2003 reads it ok?
RR handles cards formatted (ie. PARTITIONED) by just about
every device I could find (except for humans living in the dark
ages;) ... don't have any of those on hand;))))
Actually all environment I tryed this CF in read and write it.
Well, as I suspected, it's not that much Ernest's fault. Today I got
report of similar failure, with following in syslog:
--------
mmcblk0: mmc0:ff94 SD512 500224KiB
mmcblk0: unknown partition table
--------
I was told that this card was formated with WinXP. Proceeding to
investigate, I found following: if a flash card without existing
partition table is being formatted by WinXP, it will be formatted as
bare filesystem, without a partition table.
Blame first, it later become understood why WinXP behaves this way -
it simply thinks that removable devices don't need partitions. We
didn't partition diskettes, did we? Well, WinXP thinks we won't the flash
as well. So it doesn't allow to create a partition on a flash
card/drive (only delete), and formats a bare flash prtitionless
(thanks god it preserves partition table if it already exists).
This basicly means that a common user with a WinXP (I bet, any
Windows) box and w/o a camera handy, simply cannot format card in a
way suitable for RR.
And btw, partitionless storage is not just Windows' whim,
googling "partitionless flash" gives idea that it is actually in use.
I immediately proceeded to see if we can easily fix that. My
plan was (parted):
1. Resize this single partition to make a room for partition table
2. mklable msdos
3. mkpart partition, hoping that that will *add* existing partition to
the table, not harming it (partition) in any way.
Well, that kinda worked on empty card, but of course, that's of little
use. But when trying that on non-empty card, I faced parted in its
entire beauty. It seems, most common phrases it knows are "Error: Unable to
satisfy all constraints on the partition." and "No Implementation:
GNU Parted cannot resize this partition to this size. We're working
on it!". Looking at the source, there're a pretty involved machinery
happens behind the scenes (like propogating partition constraints, or
calculating cluster sizes), but those sucky messages is everything it
bothers to share with users. Btw, issue with deleting EXT2
partition/resizing FAT I experience is likely due to it throwing this
kind of errors too (didn't check it still).
So, I have following ideas: the issue seems to be pretty grave imho,
to postpone 0.6.4 and at least detect such card layout, and tell user
what's guilty and how get around the situation.
More forward-looking plan is to upgarde to parted 1.7 and patch it
too print some useful info on such failures, if they are not fixed in
it still.
_______________________________________________
Hx4700-port mailing list
[email protected]
https://www.handhelds.org/mailman/listinfo/hx4700-port
_______________________________________________
Hx4700-port mailing list
[email protected]
https://www.handhelds.org/mailman/listinfo/hx4700-port
