Re: dev/vx/if_vx_pci.c 5.0-CURRENT
On Wed, 5 Apr 2000, Eric D. Futch wrote: Ok.. I figured I'd take this on on my way to getting more comfy with FreeBSD. Check out my patch and let me know how far I've gone off the deep end :) http://quake.nyct.net/~efutch/FreeBSD/if_vx_pci.c.patch Unfortunately I do not have the acutal card to test this patch. It did survive a buildkernel, so I'm just assuming that I'm not that far off. I've actually ignored the 'vx' driver since its functionality will be folded into the 'ep' driver at some point in the future. I'm not sure its worth messing with it. -- | Matthew N. Dodd | '78 Datsun 280Z | '75 Volvo 164E | FreeBSD/NetBSD | | [EMAIL PROTECTED] | 2 x '84 Volvo 245DL| ix86,sparc,pmax | | http://www.jurai.net/~winter | This Space For Rent | ISO8802.5 4ever | To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
You really want to look at other drivers, how they have been done. When you are done, send me the diff and I will have a look at it. It is fairly straightforward to convert from one to the other, once you have understood both concepts a bit. Nick On Wed, 5 Apr 2000, Eric D. Futch wrote: I figured it would take more work than I put into it. It was just a shot in the dark. I think I have one of these cars in my Linux box at home which I wouldn't mind ripping out and sticking in this machine. Let me know if it's still worth me poking around in, or if you're going ahead with the rewrite. I'm gonna keep playing with it out of my own interest on newbusifying things. Maybe once I get a bigger clue of what has to be done, I'll see about newbusifying other things. I always have all this extra time on my hands :) -- Eric Futch New York Connect.Net, Ltd. [EMAIL PROTECTED] Technical Support Staff http://www.nyct.net (212) 293-2620 "Bringing New York The Internet Access It Deserves" On Wed, 5 Apr 2000, Brooks Davis wrote: On Wed, Apr 05, 2000 at 08:43:30PM -0400, Eric D. Futch wrote: I think if you install the kernel and reboot it will likely explode in probe. It takes more then just new structs. The function signatures have changed as well. Take a look at rev 1.67 of src/sys/pci/if_fxp.c to see what newbus conversion takes. I'm taking a wack at a rewrite, but like you I don't have a card, so testing will be kinda trickey. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message -- [EMAIL PROTECTED] [EMAIL PROTECTED] USB project http://www.etla.net/~n_hibma/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
On Thu, Apr 06, 2000 at 12:00:53AM +0200, Stephan van Beerschoten wrote: Maybe there are some people with the same problem .. or maybe someone who knows what this could be and how this problem sneaked into the system. I'm cunning -CURRENT for ages but since March 19 I am unable to compile myself a new kernel. I'm cunning 5.0-CURRENT but somehow I don't think that matters this time since the modification timestamp for the the vx devicedriver is Sat Jan 29 14:50:32 2000 (cvs timestamp that is). Its about the following ethernetcard: (info from dmesg of running kernel) vx0: 3COM 3C590 Etherlink III PCI port 0xec00-0xec1f irq 11 at device 17.0 on pci0 With a nicely updated cvs sourctree it fails on me when I run make. I get the following warnings and the final error: See /usr/src/UPDATING entry dated 2319. It looks like you are missing the required "option COMPAT_OLDPCI". The PCI portion of the vx driver needs to be converted to newbus. It looks pretty straight forward if you're feeling a moderatly ambitious. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
Ok.. I figured I'd take this on on my way to getting more comfy with FreeBSD. Check out my patch and let me know how far I've gone off the deep end :) http://quake.nyct.net/~efutch/FreeBSD/if_vx_pci.c.patch Unfortunately I do not have the acutal card to test this patch. It did survive a buildkernel, so I'm just assuming that I'm not that far off. -- Eric Futch New York Connect.Net, Ltd. [EMAIL PROTECTED] Technical Support Staff http://www.nyct.net (212) 293-2620 "Bringing New York The Internet Access It Deserves" On Wed, 5 Apr 2000, Brooks Davis wrote: See /usr/src/UPDATING entry dated 2319. It looks like you are missing the required "option COMPAT_OLDPCI". The PCI portion of the vx driver needs to be converted to newbus. It looks pretty straight forward if you're feeling a moderatly ambitious. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
On Wed, Apr 05, 2000 at 08:43:30PM -0400, Eric D. Futch wrote: Ok.. I figured I'd take this on on my way to getting more comfy with FreeBSD. Check out my patch and let me know how far I've gone off the deep end :) http://quake.nyct.net/~efutch/FreeBSD/if_vx_pci.c.patch Unfortunately I do not have the acutal card to test this patch. It did survive a buildkernel, so I'm just assuming that I'm not that far off. I think if you install the kernel and reboot it will likely explode in probe. It takes more then just new structs. The function signatures have changed as well. Take a look at rev 1.67 of src/sys/pci/if_fxp.c to see what newbus conversion takes. I'm taking a wack at a rewrite, but like you I don't have a card, so testing will be kinda trickey. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
On Wed, Apr 05, 2000 at 08:43:30PM -0400, Eric D. Futch wrote: Ok.. I figured I'd take this on on my way to getting more comfy with FreeBSD. Check out my patch and let me know how far I've gone off the deep end :) http://quake.nyct.net/~efutch/FreeBSD/if_vx_pci.c.patch Unfortunately I do not have the acutal card to test this patch. It did survive a buildkernel, so I'm just assuming that I'm not that far off. Take a look at: http://www.one-eyed-alien.net/~brooks/FreeBSD/if_vx.diff It does a newbus coversion and makes the pci probe function look half way normal. It compiles without COMPAT_OLDPCI and the kernel boots so the probe function is probably ok. I'm less sure about the attach function. This one is modeled after the eisa attach function. Testing would be appreciated, but don't try this on a production system. ;-) This driver needs a lot more work. It has a very dated view of the world and should be updated. Maybe I'll pick up one of these cards on Ebay and take a shot at it. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
I figured it would take more work than I put into it. It was just a shot in the dark. I think I have one of these cars in my Linux box at home which I wouldn't mind ripping out and sticking in this machine. Let me know if it's still worth me poking around in, or if you're going ahead with the rewrite. I'm gonna keep playing with it out of my own interest on newbusifying things. Maybe once I get a bigger clue of what has to be done, I'll see about newbusifying other things. I always have all this extra time on my hands :) -- Eric Futch New York Connect.Net, Ltd. [EMAIL PROTECTED] Technical Support Staff http://www.nyct.net (212) 293-2620 "Bringing New York The Internet Access It Deserves" On Wed, 5 Apr 2000, Brooks Davis wrote: On Wed, Apr 05, 2000 at 08:43:30PM -0400, Eric D. Futch wrote: I think if you install the kernel and reboot it will likely explode in probe. It takes more then just new structs. The function signatures have changed as well. Take a look at rev 1.67 of src/sys/pci/if_fxp.c to see what newbus conversion takes. I'm taking a wack at a rewrite, but like you I don't have a card, so testing will be kinda trickey. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
On Wed, Apr 05, 2000 at 09:35:31PM -0400, Eric D. Futch wrote: I figured it would take more work than I put into it. It was just a shot in the dark. I think I have one of these cars in my Linux box at home which I wouldn't mind ripping out and sticking in this machine. Let me know if it's still worth me poking around in, or if you're going ahead with the rewrite. I'm gonna keep playing with it out of my own interest on newbusifying things. Maybe once I get a bigger clue of what has to be done, I'll see about newbusifying other things. I always have all this extra time on my hands :) I've posted a rewrite, but I can't test it. You might go ahead and take the card out of the Linux box and play around with it. There's plenty of other work to be done on this driver to make it fully modern beyond just a simple newbus conversion. The current static storage of softc's really should be replaced. A bus_space conversion would also be a good thing. Storing the resource pointers in the softc for removal in the detach function would probalby be a good thing. If you take a look at PR kern/17601 you'll find a patch I wrote to modernize the tx driver. In and of it self, it's probably too complicated to be very useful to learn from, but I mention four revisions of if_fxp which do individual parts of the conversion. Upgrading tx was a good six hourish exercise. I've got a much better understanding of probe and attach now. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: dev/vx/if_vx_pci.c 5.0-CURRENT
In message [EMAIL PROTECTED] Brooks Davis writes: : http://www.one-eyed-alien.net/~brooks/FreeBSD/if_vx.diff +irq = bus_alloc_resource(dev, SYS_RES_IRQ, rid, + 0, ~0, 1, RF_ACTIVE); irq = bus_alloc_resource(dev, SYS_RES_IRQ, rid, 0, ~0, 1, RF_ACTIVE | RF_SHAREABLE); pci devices can share interrupts. Otherwise I didn't see anything wrong with this. I have a 905B, so I can't test this driver on hardware I have :-( Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
