Messages in this thread Patch in this message | | | Date | Tue, 21 Sep 1999 13:06:05 +0200 | From | Petr Vandrovec <> | Subject | Re: current 2.3.18ac & debian-potato's gcc & something? |
| |
On 21 Sep 99 at 11:13, Alan Cox wrote: > > And another strange thing, ac6 is not able to correctly checksum > > incoming tcp/udp frames (TCPv4 bad checksum .., UDP: > > bad checksum, 100% fails or I missed someone correctly checksummed) > Some work is going on experimenting with using the various memory rather > than I/O operations for some net cards. de4x5 suffers from same problem under my ac6 :-( And problem is limited to IP only, because of I downloaded patch-2.3.18ac7.bz2 on my ac6 box through IPX (ncpfs) and file came correct. Also icmp ping does not have any troubles. > That the tulip wasnt detected in ac7 is odd. Either all pci_id_info tables in network drivers are wrong or pci_drv_register swaps vendor/device fields. I fixed only minimum in drivers/pci/scan.c. Maybe that subsystem_vendor/device should be swapped too. But if I could vote, I'm voting for swapping vendor/device in structures, 0x00191011 does not mean anything to me, but 0x10110019 does, instead of changing pci_drv_register. But there are too much structures, but only one pci_drv_register...
diff -urdN linux/drivers/pci/scan.c linux/drivers/pci/scan.c --- linux/drivers/pci/scan.c Tue Sep 21 08:35:55 1999 +++ linux/drivers/pci/scan.c Tue Sep 21 10:41:09 1999 @@ -96,7 +96,7 @@ unsigned long ioaddr; int free = 0; - pci_id = (pdev->vendor << 16) | pdev->device; + pci_id = (pdev->device << 16) | pdev->vendor; pci_subsys_id = (pdev->subsystem_vendor << 16) | pdev->subsystem_device; pci_read_config_byte (pdev, PCI_REVISION_ID, &pci_rev); > That it got frames wrong is not tulip now works under ac7... I also changed MMIO region size from 1KB to 128 bytes as my 10110019 has only 128 bytes MMIO (and IO too). Best regards, Petr Vandrovec vandrove@vc.cvut.cz
P.S.: I do not know, what was wrong with ac6...
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |