Am Dienstag 31 Oktober 2006 11:40 schrieb Christoph Hellwig: > On Tue, Oct 31, 2006 at 04:43:19PM +0800, Peer Chen wrote: > > + u32 cb_add, temp32; > > + struct device *dev = pci_dev_to_dev(pdev); > > + struct ata_host_set *host_set = dev_get_drvdata(dev); > > + u8 pro=0; > > + if (!(pro&0x40)) > > + return; > > + > > + temp32 = csr_add; > > + phost->host_sgpio.pcsr = (void *)temp32; > > + phost->host_sgpio.pcb = phys_to_virt(cb_add); > > Use of phys_to_virt is generally a bug. What are you trying to do here? I am also wondering whether casting of temp32 to a pointer is very 64bit friendly? At least my compiler on x86_64 gives a warning... I also found http://lkml.org/lkml/2006/8/21/324 which looks alike at first glance and it seems NVidia didn't really fix what Andrew Morton told them to do so... Cheers, -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V
Attachment:
pgpLdikL3TanO.pgp
Description: PGP signature
