On Sun, 14 Sep 2008 01:48:06 +0400, Sergei Shtylyov <sshtylyov@xxxxxxxxxxxxx> wrote:
> > +#ifdef __BIG_ENDIAN
> > + /* This setting does not affect PRD fetch */
> > + /* ByteSwap=1, Endian=00 */
> > + TX4939IDE_writew(0xc911, base, Add_Ctl);
> > +#else
> > + TX4939IDE_writew(0xc901, base, Add_Ctl);
> > +#endif
> >
>
> Aren't these the default register values? Is there a sense in writing
> them?
Indeed. It seems redundant.
> > +#ifdef __BIG_ENDIAN
> > +/* custom iops (independent from SWAP_IO_SPACE) */
> > +static u8 mm_inb(unsigned long port)
> > +{
> > + return (u8)readb((void __iomem *)port);
> > +}
> > +static void mm_outb(u8 value, unsigned long port)
> > +{
> > + writeb(value, (void __iomem *)port);
> > +}
> > +static void mm_tf_load(ide_drive_t *drive, ide_task_t *task)
> > +{
> > + ide_hwif_t *hwif = drive->hwif;
> > + struct ide_io_ports *io_ports = &hwif->io_ports;
> > + struct ide_taskfile *tf = &task->tf;
> > + u8 HIHI = (task->tf_flags & IDE_TFLAG_LBA48) ? 0xE0 : 0xEF;
> > +
> > + if (task->tf_flags & IDE_TFLAG_FLAGGED)
> > + HIHI = 0xFF;
> > +
> > + if (task->tf_flags & IDE_TFLAG_OUT_DATA) {
> > + u16 data = (tf->hob_data << 8) | tf->data;
> > +
> > + __raw_writew(data, (void __iomem *)io_ports->data_addr);
> >
>
> This doesn't look consistent (aside from the TX4939IDE_REG8/16 issue)
> -- mm_outsw_swap() calls cpu_to_le16() before writing 16-bit data but
> this code doesn't. So, either one of those should be wrong...
Thanks, this code should be wrong. IDE_TFLAG_OUT_DATA is totally
untested...
> > +static void mm_insw_swap(unsigned long port, void *addr, u32 count)
> > +{
> > + unsigned short *ptr = addr;
> > + unsigned long size = count * 2;
> > + port &= ~1;
> > + while (count--)
> > + *ptr++ = le16_to_cpu(__raw_readw((void __iomem *)port));
> > + __ide_flush_dcache_range((unsigned long)addr, size);
> > +}
> > +static void mm_outsw_swap(unsigned long port, void *addr, u32 count)
> > +{
> > + unsigned short *ptr = addr;
> > + unsigned long size = count * 2;
> > + port &= ~1;
> > + while (count--) {
> > + __raw_writew(cpu_to_le16(*ptr), (void __iomem *)port);
> > + ptr++;
> > + }
> > + __ide_flush_dcache_range((unsigned long)addr, size);
> > +}
> >
>
> Hum... but is it really correct to convert from/to LE order above?
> I'm prett sure that data is expected in LE order -- look ar
> ide_fix_driveid() for example...
Well, do you mean I should use cpu_to_le16 in mm_insw_swap and
le16_to_cpu in mm_outsw_swap? Or can I avoid these swapping entirely
in some way?
---
Atsushi Nemoto
--
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
