From mboxrd@z Thu Jan 1 00:00:00 1970 From: "David S. Miller" Date: Sun, 18 May 2003 09:35:33 +0000 Subject: [Linux-ia64] Re: [patch] support 64 bit pci_alloc_consistent Message-Id: List-Id: References: In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org From: Arjan van de Ven Date: 18 May 2003 11:29:02 +0200 #define PCI_DMA_64BIT 0xffffffffffffffffULL #define PCI_DMA_32BIT 0xffffffffULL void pci_set_dma_capabilities(device, u64 streaming_mask, u64 persistent_mask); u64 pci_get_effective_streaming_mask(device); u64 pci_get_effective_persistent_mask(device); if for some reason the architecture PCI code needs or wants to reduce the DMA mask WHile logically you are correct, the probing code is going to look basically identical. Instead of frobbing around with pci_set_dma() calls, you're going to be frobbing around with pci_get_effective*() calls and branching based upon that. I really see no value in this. Both are effectively equivalent and the present setup has the advantage that it has existed for some time and driver authors (at least some :-) know how to use it already.