Re: Proposal for a new PCI function call

2001-04-18 Thread Jes Sorensen
> "Alan" == Alan Cox <[EMAIL PROTECTED]> writes: >> Introducing a new function that takes bit flags as arguments might >> be better? Alan> pci_set_dma_mask_bits() ? So you could do Alan> pci_set_dma_mask_bits(pdev, 64); Alan> We want everything to go through pci_set_dma_mask... type Alan>

Re: Proposal for a new PCI function call

2001-04-18 Thread Jes Sorensen
> "Jeff" == Jeff Garzik <[EMAIL PROTECTED]> writes: Jeff> Jes Sorensen wrote: >> Hmmm, I was wondering if could come up with a pretty way to do this >> on 32 bit boxes that wants to enable highmem DMA. Right now >> pci_set_dma_mask() wants a dma_addr_t which means you have to do >> #ifdef CON

Re: Proposal for a new PCI function call

2001-04-16 Thread Venkatesh Ramamurthy
> It seems to me that not doing #ifdef CONFIG_HIGHMEM right now is a > bug... I think it's the megaraid driver that wants to set dma_addr_t to > a 64-bit mask. MegaRAID driver: Only if the flag __LP64__ is defined, a 64 bit mask is set , otherwise only a 32 bit mask is used instead. However chec

Re: Proposal for a new PCI function call

2001-04-13 Thread Jeff Garzik
Jes Sorensen wrote: > > "Jeff" == Jeff Garzik <[EMAIL PROTECTED]> writes: > >> I think the function idea would let us do some sanity checking to > >> make sure drivers weren't setting this to 64bit on non-64 bit > >> busses and stuff. > Jeff> pci_set_dma_mask. Modify that to do the additiona

Re: Proposal for a new PCI function call

2001-04-12 Thread Alan Cox
> Hmmm, I was wondering if could come up with a pretty way to do this on > 32 bit boxes that wants to enable highmem DMA. Right now > pci_set_dma_mask() wants a dma_addr_t which means you have to do > #ifdef CONFIG_HIGHMEM #else #endif. > > Introducing a new function that takes bit flags as arg

Re: Proposal for a new PCI function call

2001-04-12 Thread Jes Sorensen
> "Jeff" == Jeff Garzik <[EMAIL PROTECTED]> writes: >> I think the function idea would let us do some sanity checking to >> make sure drivers weren't setting this to 64bit on non-64 bit >> busses and stuff. Jeff> pci_set_dma_mask. Modify that to do the additional checks you Jeff> need. Jef

Re: Proposal for a new PCI function call

2001-04-12 Thread Jeff Garzik
Steve Modica wrote: > > Hi All, > > We found recently that the acenic driver for the 3com gigabit ethernet card does > not enable 64 bit DMAs. (this is done by setting the appropriate mask in > pci_dev->dma_mask). > > Jes suggested that the appropriate way to fix this would be to create a func