[openib-general] 2.6.18-rc5-mm1: drivers/infiniband/hw/amso1100/c2.c compile error

Andrew Morton akpm at osdl.org
Fri Sep 1 11:23:12 PDT 2006


On Fri, 01 Sep 2006 10:34:24 -0700
Roland Dreier <rdreier at cisco.com> wrote:

>     Andrew> What's __raw_writeq() supposed to do, anyway?  On alpha
>     Andrew> it's writeq() without an mb().  On parisc it's writeq()
>     Andrew> only the data is byte-reversed.  On sparc64() it's
>     Andrew> incomprehensible.  On everything else it's writeq().
> 
> My understanding is that __raw_writeq() is like writeq() except not
> strongly ordered and without the byte-swap on big-endian
> architectures.  The __raw_writeX() variants are convenient to avoid
> having to write inefficient code like writel(swab32(foo), ...) when
> talking to a PCI device that wants big-endian data.  Without the raw
> variant, you end up with a double swap on big-endian architectures.
> 
> sparc64 looks wrong, since __raw_writeq() seems identical to writeq(),
> which seems to imply it's going to swab what is stores.
> 

OK.  Can we please stop hacking around this in drivers and

a) work out what it's supposed to do

b) document that (Documentation/DocBook/deviceiobook.tmpl or code
   comment or whatever)

c) tell arch maintainers?




More information about the general mailing list