On Wed, 28 Oct 2009 13:54:53 +0800
joy zhao <yanhwizhao@xxxxxxxxx> wrote:
> hi，
> after read about pci iomap code, I found some inconsistence in
> function pci_iomap() for x86 platform.
> I don't know whether it's indeed an inconsistence or an intended
> design, or somewhere I'm wrong?
>
> see:
>
> pci_iomap(), for x86 platform, is defined in
> include/asm-generic/iomap.h and it has a implementation in
> lib/iomap.c, that is,
>
>
> void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long
> maxlen) {
> resource_size_t start = pci_resource_start(dev, bar);
> resource_size_t len = pci_resource_len(dev, bar);
> unsigned long flags = pci_resource_flags(dev, bar);
>
> if (!len || !start)
> return NULL;
> if (maxlen && len > maxlen)
> len = maxlen;
> if (flags & IORESOURCE_IO)
> return ioport_map(start, len);
> if (flags & IORESOURCE_MEM) {
> if (flags & IORESOURCE_CACHEABLE)
> return ioremap(start, len);
> return ioremap_nocache(start, len);
> }
> /* What? */
> return NULL;
> }
>
>
> from which, we can see,
> if (flags & IORESOURCE_CACHEABLE)
> return ioremap(start, len);
>
> but, after look into ioremap, i found its x86 implementation(in
> arch/x86/include/asm/io.h) is
>
> static inline void __iomem *ioremap(resource_size_t offset, unsigned
> long size) {
> return ioremap_nocache(offset, size);
> }
>
> so, ioremap is not the cached version as expected.
Hm, yeah seems like pci_iomap should call ioremap_cache() in the
IORESOURCE_CACHEABLE case now that we have the various ioremap
variants... Venki?
--
Jesse Barnes, Intel Open Source Technology Center
--
To unsubscribe from this list: send the line "unsubscribe linux-pci" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
