PDA

Bekijk de volledige versie : FDISK and other utils wanted



Technik
08-01-2004, 14:51
I am trying to make the USB flash drive APACER Handy Steno SL202 working with WL-500b. This device should work with Linux Kernel 2.4.0 or above and in the manual is a description how to initialize the flash drive. Therefore I need at least the FDISK and MKFS.VFAT utility. Unfortunately I did not find them compiled for MIPS anywhere so far - only source codes found which I am not able to compile yet. Although the FDISK is described in BUSYBOX documentation, the FDISK is not included in any BUSYBOX bin I have. Could anybody help me please?

FIB
08-01-2004, 18:46
I have formatted the [USB] Memory Sticks with a Microsoft XP, both as FAT and FAT32, and afterwards used it on the WL500g without problems. I have not figured out how to create several partitions yet.

But are you sure that the APACER Handy Steno SL202 can work with Wl500g? I got problmes with the older version of SL202! See Reported Compatible Hardware -> [USB] Memory Sticks in this forum.

Technik
08-01-2004, 21:50
Originally posted by FIB
I have formatted the [USB] Memory Sticks with a Microsoft XP, both as FAT and FAT32, and afterwards used it on the WL500g without problems.
I tried the same under W2K as well already - it did not help of course. I am afraid formatting from Linux will not fix the problem but sometimes it could be fine to have a possibility to analyze/initialize a flash disk filesystem with FDISK or similar utility.



But are you sure that the APACER Handy Steno SL202 can work with Wl500g? I got problmes with the older version of SL202! See Reported Compatible Hardware -> [USB] Memory Sticks in this forum.
I know about that and I think it could be really a HW problem (or Linux USB driver problem) but I would like to follow APACER's instructions first. My USB key version seems to be even newer - the serial number ends with b3.

Oleg
09-01-2004, 10:45
It's possible to compile fdisk in (not sure for mkfs.vfat), but it's better to try this on the linux box first to be completely sure, that this helps. In my opinion, we should not compile the utils like this to busybox, cause your always can use windows or linux box for the purpose and not waste the flash space. I'm using CF card inserted to the USB reader and formated by windows. Try using dos format utility also.

zomk3
09-01-2004, 11:14
Originally posted by Oleg
It's possible to compile fdisk in (not sure for mkfs.vfat), but it's better to try this on the linux box first to be completely sure, that this helps. In my opinion, we should not compile the utils like this to busybox, cause your always can use windows or linux box for the purpose and not waste the flash space. I'm using CF card inserted to the USB reader and formated by windows. Try using dos format utility also.

I agree with oleg. It way easier to do this on a usual linux box. Its better to use the space in the flash for things like vpn deamons aso. You usually only need tools like this once. It would say that only things needed often by the router should go in the custom firmware. After you partitioned the stick, you can exensivly store other software on it. But remember the box has only 16 megs of Ram.

Technik
09-01-2004, 11:47
Well, I really did not mean to put the compiled "full" busybox to router's flash at all. I would just like to have a "full" busybox binary which I could copy to /tmp folder temporarily and run a requested utility one time.

Technik
09-01-2004, 11:56
Originally posted by zomk3
But remember the box has only 16 megs of Ram.
Yes I know - it's a really annoying limit. :( But what about the nice undocumented expansion slot on the mainboard? ;) I don't know how much memory can the processor address (and if the hardware design could ever allow to add more memory) but I think at least 64 MB of RAM would be much better... ;)

Oleg
09-01-2004, 12:01
ok, just wait. You will got the busybox 1.0-pre5 with all things compiled in and statically linked against glibc. But it will be really big. Just like the firmware. ;-)

Oleg
09-01-2004, 12:31
Done it, and send to Antiloop. This binary has also "stand-alone shell" & "applets always win" features, which means that once you've started the shell from this busybox binary you will get everything, just like it was installed in the flash. Do it like this
"/mnt/usbfs/busybox sh" from the shell prompt.

Technik
09-01-2004, 12:45
Excellent, Oleg, thanks a lot again!!! :) I hope some other users will take an advantage of this busybox as well.

Antiloop
09-01-2004, 18:04
Originally posted by Oleg
Done it, and send to Antiloop. This binary has also "stand-alone shell" & "applets always win" features, which means that once you've started the shell from this busybox binary you will get everything, just like it was installed in the flash. Do it like this
"/mnt/usbfs/busybox sh" from the shell prompt.


file can be found here:

http://files.wl500g.info/asus/wl500g/firmware/app/busybox.zip

zomk3
09-01-2004, 21:05
Originally posted by Technik
Yes I know - it's a really annoying limit. :( But what about the nice undocumented expansion slot on the mainboard? ;) I don't know how much memory can the processor address (and if the hardware design could ever allow to add more memory) but I think at least 64 MB of RAM would be much better... ;)
I am not sure if i understand the broadcom sources right, but it seems as if it can address 128MB of RAM.

Did i miss something about an expansion slot ?

I think in "older" days on my amiga there was a solution for a problem like this. You could mould the new over the old rams and just connect one adress pin directly to the processor thru a cmos gate...

Technik
12-01-2004, 14:13
Originally posted by zomk3
I am not sure if i understand the broadcom sources right, but it seems as if it can address 128MB of RAM.

Did i miss something about an expansion slot ?

I think in "older" days on my amiga there was a solution for a problem like this. You could mould the new over the old rams and just connect one adress pin directly to the processor thru a cmos gate...

Well, 128MB would be nice. :)

I saw some connector on mainboard picture - maybe there's a way to increase RAM capacity using that connector. I did not open my own unit yet (I am aware of warranty a bit) so I did not analyze it so far.

Technik
13-01-2004, 09:36
It seems that APACER Handy Steno SL202 can't work with current version of USB driver - it is recognized as SCSI device but there is an error during the communication so dev/sda is not created. Therefore the FDISK can't do anything... It will be better to replace APACER with PQI USB flash disk...

Oleg
13-01-2004, 16:32
Just to warn you - PQI write performance is very poor. Apacer flashes works faster. Transcend ones also fast enough.

Technik
13-01-2004, 16:57
OK, Oleg, thanks for the warning. Maybe APACER is too fast to work properly with WL-500b/g. I prefer the reliability over speed anway so I hope the PQI selection (based on Antiloop's report in RCH forum) was not a big mistake. If the performance will be too slow, I will return it to supplier again... ;)

Antiloop
13-01-2004, 20:14
Originally posted by Technik
OK, Oleg, thanks for the warning. Maybe APACER is too fast to work properly with WL-500b/g. I prefer the reliability over speed anway so I hope the PQI selection (based on Antiloop's report in RCH forum) was not a big mistake. If the performance will be too slow, I will return it to supplier again... ;)

hehe it's a usb2.0 device.. so speed won't be slow i guess..
never really tested..

how much bandwidth has usb1.1 anyway?

Oleg
14-01-2004, 10:06
The problem with PQI write performance is not the interface speed, but the flash speed. That's the russian review of the drives - http://www.ixbt.com/storage/flashdrives-p9-canyon-jetlinx-kingston-ms-pqi-transcend-xdrive.shtml
There are 3 graphs at the end - the first one is the access time in ms (lower is better), the next one is the read performance (higher is better) and the later is the write performance (KBytes/s).
All are using USB 2.0.
As for USB 1.1 - the bandwidth is something near 1 Mbytes/s.

Technik
14-01-2004, 11:43
Well, I think the flash disk speed can also depend on the flash capacity. There could be also some hardware design changes between various models of the same manufacturer so we should not judge the manufacturer pursuant to one specific model test. We should compare all product ranges I think but it's not possible to test all devices over the world... Anyway, thanks for the link to test - it's interresting to see the differencies and it can help to somebody who want to buy some of listed models.