[Zaurus-devel] pxa pcmcia timings

2010-09-02 Thread kirill
Dear zaurus developers! I used to delve into zaurus code for awhile ago. There are some local changes left from that time. One thing is the change to pxa pcmcia timings. Here's the original code from 2.6.16: static inline u_int pxa2xx_mcxx_hold(u_int pcmcia_cycle_ns,

[Zaurus-devel] LCD dma turning

2010-09-02 Thread kirill
To reduce LCD dma interruption (i.e. LCD artifacts during pcmcia transfers) we can raise it's priority by means of LCSBSCNTR and ARB_CNTRL registers. I used the following values on c3000: LCDBSCNTR = 2 ARB_CNTRL = 0x01800f11 Those registers can be set in spitc.c/common_init, but I guess that the

Re: [Zaurus-devel] LCD dma turning

2010-09-02 Thread kirill
kirill wrote: > To reduce LCD dma interruption (i.e. LCD artifacts during pcmcia transfers) > we can raise it's priority > by means of LCSBSCNTR and ARB_CNTRL registers. > I used the following values on c3000: > LCDBSCNTR = 2 > ARB_CNTRL = 0x01800f11 > > Those re

Re: [Zaurus-devel] pxa pcmcia timings

2010-09-06 Thread kirill
> Yes, patch would be nice. I am currently debugging some PCMCIA > problems... Question: is/was there situation where your version worked > but old one did not? No. > Does it provide faster data rates? Yes, it does. As I recall, pcmcia spec doesn't provide "fomula" to calculate cycle timings. It d

Re: [Zaurus-devel] LCD dma turning

2010-09-06 Thread kirill
Pavel Machek wrote: > On Thu 2010-09-02 19:01:05, kirill wrote: >> kirill wrote: >>> To reduce LCD dma interruption (i.e. LCD artifacts during pcmcia transfers) >>> we can raise it's priority >>> by means of LCSBSCNTR and ARB_CNTRL registers.

Re: [Zaurus-devel] pxa pcmcia timings

2010-09-10 Thread kirill
kirill wrote: >> Yes, patch would be nice. Here's the patch. I didn't test it on 2.6.35, however the affected functions are not changed from 2.6.1x. So it should work... Note pcmcia_cycle_ns in original code. AFAIK it's not pcmcia cyce ns. See my FIXME below. diff -ru linu

[Zaurus-devel] cf speed evolution

2010-09-21 Thread kirill
Hi! Recently I've upgraded internal cf on my c3000. So now I have a better idea of what cf storage performance you can get with c3000 hardwire. Benchmark results (2.6.19 kernel, dd if=/dev/hda of=/dev/null): 1) unmodified kernel: ~ 2MB/sec 2) optimized timings, 250ns io cycle: ~ 3MB/sec 3) optim

Re: [Zaurus-devel] cf speed evolution

2010-09-26 Thread kirill
Well, I tried your patch from the list on 2.6.26-RP. I see 2.33 MB/sec -> 2.63MB/sec improvement on 6 years old 40x card. In my experience, the speed in pcmcia mode is not related to card's maximum speed. I don't think that 6 year old CF supports advanced timing mode. You may check it with hdp

Re: [Zaurus-devel] cf speed evolution - ide-cs VDMA patch

2010-09-27 Thread kirill
19:46:44.0 +0400 @@ -0,0 +1,430 @@ +/* + * Support for "virtual" DMA transfers using the DMA controller of pxa2xx processors. + * Currently this code is only tested on PXA270 processor. + * + * Copyright 2006 Kirill Kornilov + * + */ + +#include +#include +#include +#include +#inc