Hi,
It's been a while since I deactivated UDMA on my micro-A1 due to data corruption problems, but since no-one seems to report the same issues (on the micro I mean), I may have missed something (I'm quite sure that UDMA did work at some time though).
So, when I activate even the slowest UDMA mode on my hard disk or my CD burner, strange things start to happen. Workbench does not list all icons in drawers, programs randomly crash, and eventually the system hangs. On the CD drive side, if I rip a song I can hear a periodical "clic" noise when I listen to the resulting mp3. Actually, I tracked the problem further when I was debugging ADRipper. The first 32 data bytes of every IDE transaction are overwritten by a series of 0xFF. I assume the same kind of thing happens with the hard disk, which would explain the random crashes.
I am using a 80-wire IDE cable (tried with both a flat ribbon and a round one). Due to SDRAM issues the FSB of my micro is underclocked at 100 MHz.
So, is this some kind of known VIA686B bug or could I have made a configuration mistake somewhere ?
Thanks for help,
CentaurZ
UDMA on microA1
-
- Beta Tester
- Posts: 233
- Joined: Sun Jun 19, 2011 5:26 pm
Re: UDMA on microA1
With my µAOne i use a bootpartition with SFS0, my second partition is formatted with SFS2.
max. transfer = 1FE00
Mask = FFFFFFFE
Blocksize: 512
Puffer: 600
My hardrive uses UDMA5 and DVD is UDMA4. Works perfectly.
With my Aone XE G4, SII0680 are all the same but max. transfer is 7FFFFFFF.
Perhaps it would help.
max. transfer = 1FE00
Mask = FFFFFFFE
Blocksize: 512
Puffer: 600
My hardrive uses UDMA5 and DVD is UDMA4. Works perfectly.
With my Aone XE G4, SII0680 are all the same but max. transfer is 7FFFFFFF.
Perhaps it would help.
Re: UDMA on microA1
I am using the same settings as yours, except that all my partitions but the boot one are JXFS/04 now. I had the same issues with SFS before so it seems to be low level (device) related.
Regarded Uboot vars, I erased all a1ide_xxx stuff so everything relies on autodetection.
Regarded Uboot vars, I erased all a1ide_xxx stuff so everything relies on autodetection.
-
- Beta Tester
- Posts: 233
- Joined: Sun Jun 19, 2011 5:26 pm
Re: UDMA on microA1
You could try this in uboot.
setenv a1ide_xfer EEEE
saveenv
------------------------------------------------------------------
0 - Automatic PIO (driver will use the best PIO mode
that the drive claims to support)
a - PIO 0 ( 3 MB/s, modeid 8)
b - PIO 1 ( 5 MB/s, modeid 9)
c - PIO 2 ( 8 MB/s, modeid 10)
d - PIO 3 ( 11 MB/s, modeid 11)
e - PIO 4 ( 16 MB/s, modeid 12)
A - UDMA 0 ( 16 MB/s, modeid 64)
B - UDMA 1 ( 25 MB/s, modeid 65)
C - UDMA 2 ( 33 MB/s, modeid 66)
D - UDMA 3 ( 44 MB/s, modeid 67)
E - UDMA 4 ( 66 MB/s, modeid 68)
F - UDMA 5 (100 MB/s, modeid 69)
setenv a1ide_xfer EEEE
saveenv
------------------------------------------------------------------
0 - Automatic PIO (driver will use the best PIO mode
that the drive claims to support)
a - PIO 0 ( 3 MB/s, modeid 8)
b - PIO 1 ( 5 MB/s, modeid 9)
c - PIO 2 ( 8 MB/s, modeid 10)
d - PIO 3 ( 11 MB/s, modeid 11)
e - PIO 4 ( 16 MB/s, modeid 12)
A - UDMA 0 ( 16 MB/s, modeid 64)
B - UDMA 1 ( 25 MB/s, modeid 65)
C - UDMA 2 ( 33 MB/s, modeid 66)
D - UDMA 3 ( 44 MB/s, modeid 67)
E - UDMA 4 ( 66 MB/s, modeid 68)
F - UDMA 5 (100 MB/s, modeid 69)
Re: UDMA on microA1
I tried all UDMA modes (with idetool), like I said even with the slowest one, no success.
- Christopher Follett
- AmigaKit
- Posts: 651
- Joined: Mon Feb 21, 2011 5:56 pm
- Location: Wales, UK
- Contact:
Re: UDMA on microA1
I had same issues, until I re-heat compound on the CPU and heat sink.
I then overclocked to 900MHz with my FSB at 100. Hardly even have any issues now, the A1 has been stable for atleast 6 months.
Ive even left it running for hourse without issue on DMA, think due to HDD, the highest I could go was DMA 3.
I then overclocked to 900MHz with my FSB at 100. Hardly even have any issues now, the A1 has been stable for atleast 6 months.
Ive even left it running for hourse without issue on DMA, think due to HDD, the highest I could go was DMA 3.
http://www.ultimateamiga.co.uk (Home of SONY PSP Amiga Emulator and AMOS Factory)Resolute and Industrious
Grand ruler of the yellow people and the Ultimate Amiga Empire
__________________
Customer Help & Support: http://www.amigakit.com/help
http://www.amigakit.com - the Amiga store
New Products
- tonyw
- AmigaOS Core Developer
- Posts: 1479
- Joined: Wed Mar 09, 2011 1:36 pm
- Location: Sydney, Australia
Re: UDMA on microA1
Don't forget the hardware needs of UDMA: you need good 80-wire flat cables, with a drive always connected on the end of the cable (never a loose flappy bit).
Don't use round cables.
Don't use old 40-wire cables.
Don't kink or fold the cable, keep it clear of other wiring.
Make sure that the cable is connected the right way around, System connector to the motherboard, the far end to the Master and the middle to the Slave.
Don't use round cables.
Don't use old 40-wire cables.
Don't kink or fold the cable, keep it clear of other wiring.
Make sure that the cable is connected the right way around, System connector to the motherboard, the far end to the Master and the middle to the Slave.
cheers
tony
tony
Re: UDMA on microA1
@Christopher Follett
That's interresting. Do you mean UDMA was causing trouble with a core frequency of 600 MHz and setting the multiplicator to 9 sort of fixed this ?
I haven't tried to overclock my G3 yet because the dip switches are glued. However, if I got it right it's worth trying. I replaced the original fan by a very good dissipator some time ago, so running 100MHz higher than nominal frequency shouldn't hurt.
That's interresting. Do you mean UDMA was causing trouble with a core frequency of 600 MHz and setting the multiplicator to 9 sort of fixed this ?
I haven't tried to overclock my G3 yet because the dip switches are glued. However, if I got it right it's worth trying. I replaced the original fan by a very good dissipator some time ago, so running 100MHz higher than nominal frequency shouldn't hurt.
- Christopher Follett
- AmigaKit
- Posts: 651
- Joined: Mon Feb 21, 2011 5:56 pm
- Location: Wales, UK
- Contact:
Re: UDMA on microA1
Exactly what I thought, but I had to play around with clck and vcore, to get it to 900MHz.centaurz wrote:@Christopher Follett
That's interresting. Do you mean UDMA was causing trouble with a core frequency of 600 MHz and setting the multiplicator to 9 sort of fixed this ?
I haven't tried to overclock my G3 yet because the dip switches are glued. However, if I got it right it's worth trying. I replaced the original fan by a very good dissipator some time ago, so running 100MHz higher than nominal frequency shouldn't hurt.
As I said after that its very stable. I recall lots of people saying same stuff, they randomly done what they wanted at factory, screwing up voltages and under clocking cpu. Think they done this to get round hardware flaw.
http://www.ultimateamiga.co.uk (Home of SONY PSP Amiga Emulator and AMOS Factory)Resolute and Industrious
Grand ruler of the yellow people and the Ultimate Amiga Empire
__________________
Customer Help & Support: http://www.amigakit.com/help
http://www.amigakit.com - the Amiga store
New Products
Re: UDMA on microA1
I've had the same problem with my uA1 since I first purchased it in 2004. Unstable with UDMA enabled. I tried changing harddrive cables, vcore voltages etc. Never could get it to run stable with UDMA mode on. The only way I was able to solve the problem was to buy an X1000. Sorry, I know that's not helpful. My uA1 still runs in PIO mode.centaurz wrote: It's been a while since I deactivated UDMA on my micro-A1 due to data corruption problems, but since no-one seems to report the same issues (on the micro I mean), I may have missed something (I'm quite sure that UDMA did work at some time though).
Kevin - X1000 first contact / uA1