> Please Cc: x86 patches to the x86 maintainers.
Sure - I should probably have also cc'd Feng - I've forwarded it and
added Feng so it doesn't get missed.
> > +static int hsu_inited;
>
> 'initialized' is the proper English word i think.
Then we should probably run /sbin/initialize in future ;) inited is
perfectly fine computerspeak and much less typing.
> > +static void early_hsu_init(void)
> > +{
> > + u8 lcr;
> > +
> > + if (phsu && hsu_inited)
> > + return;
>
> Surely one of those will suffice as a "have we initialized" flag?
>
> Also, under what circumstances can we call early_hsu_init() twice?
Don't think we can
> > + /* GPIO workaround */
> > + set_fixmap_nocache(FIX_EARLYCON_MEM_BASE,
> > MFD_GPIO_HSU_REG);
> > + phsu = (void *)(__fix_to_virt(FIX_EARLYCON_MEM_BASE) +
> > + (MFD_GPIO_HSU_REG & (PAGE_SIZE - 1)));
> > +
> > + *((u32 *)phsu) = 0x55465;
>
> What does 0x55465 stand for?
It's a firmware fixup. Its a magic value (even to most of us who work
here ;)). Feng - am I right in thinking we don't need that anyway with
the current firmware ?
> > +{
> > + unsigned int timeout = 10000; /* 10ms*/
> > + u8 status;
> > +
> > + while (timeout--) {
> > + status = readb(phsu + UART_LSR);
> > + if (status & BOTH_EMPTY)
> > + break;
> > +
> > + udelay(1);
> > + }
> > +
> > + if (timeout == 0xffffffff)
> > + return;
>
> Using the -1 literal will dtrt too, and will be slightly clearer to
> the potentially overworked reader of such patches.
If they have a degree in C sign propagation. Better is probably
while (--timeout) {..
}
--
To unsubscribe from this list: send the line "unsubscribe linux-serial" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
