Hi,
On 11-05-15 20:53, Arnd Bergmann wrote:
On Monday 11 May 2015 20:14:12 Hans de Goede wrote:
Also note, that as I've tried to explain in the commit message already this commit
makes no functional changes what so ever, the generic musb_default_read/write_fifo
functions are only used by musb platforms which do not overwrite musb_write* /
musb_read* and musb_write* / musb_read* are already just thin wrappers around
the __raw_* functions.
Oh, that's a bit odd: That would mean that the defaults are also broken on
big-endian platforms, and fixing them would introduce the bug here, unless your
patch is applied.
Any idea why musb uses the __raw_* functions to start with? Can we change
them to read?_relaxed()/write?_relaxed() now that those are available on
all architectures?
I'm afraid I've no idea, since the musb code is somewhat obscure / tricky
in places I've done my best to not make any functional changes what so ever
in the entire set, chances are things will work fine with just readl / writel
but verifying that is hard, so I've gone for the better safe then sorry approach.
Regards,
Hans
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
