Re: kldloaded driver not called at load time
I never liked the unknown driver. I vote we kill it and let isa_probe_nomatch() deal with things. (I'll have to write isa_probe_nomatch() but I don't see that as a problem.) I'll dig out my patches for that area and send you a copy. We do need something when we use PNPBIOS because the motherboard devices are not programmable and their port space is "fixed". Their space needs to either be allocated to something or otherwise avoided when we are choosing ports for real programmable devices. I was looking at this the other day, trying to work out how to differentiate easily between things that we _know_ are in a single, immovable place (eg. most of the stuff that we see from the PnP BIOS), things that "might" be there, and things that are there and can be shuffled around. Basically we need a way of sorting the ISA bus children in a more sophisticated fashion that the isa_probe_children code currently does. -- \\ Give a man a fish, and you feed him for a day. \\ Mike Smith \\ Tell him he should learn how to fish himself, \\ [EMAIL PROTECTED] \\ and he'll hate you for a lifetime. \\ [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
kldloaded driver not called at load time
I am trying to make kldloadable driver. But none of functions are called when I kldload it. Is this a bug or I am missing something essential? Sources, dmesg, config and makefile (for module) are at http://ethereal.ru/~nms/TR/tk.tar (~40Kb) P.S. I am chasing for IBM Token Ring Adapter (unknown0:). Thanks To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Nikolai Saoukh wrote:
I am trying to make kldloadable driver. But none of functions are
called when I kldload it. Is this a bug or I am missing something
essential?
You're missing something essential.
The return values for the DEVICE_PROBE method should be 0 or negative for
priority, or positive to indicate an error.
if_tok.c:tok_probe() returns ENXIO which is in the return value you use
for if_tok_isa.c:tok_isa_probe().
You should take a good look at /sys/dev/{ex,ep} for an example of how to
accommodate PnP probes.
--
| 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: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Nikolai Saoukh wrote: THE PROBLEM is that tok_isa_probe is not called at all, when driver kldloaded. Return value is irrelevant for this case. Yea, I see what could be the problem. I'll get back to you. -- | 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: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Nikolai Saoukh wrote:
THE PROBLEM is that tok_isa_probe is not called at all,
when driver kldloaded. Return value is irrelevant for this case.
Try the attached patch.
(cd /sys/isa patch isa_common.c.patch)
--
| 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 |
Index: isa_common.c
===
RCS file: /cvs/src/sys/isa/isa_common.c,v
retrieving revision 1.15
diff -u -r1.15 isa_common.c
--- isa_common.c2000/01/18 02:15:05 1.15
+++ isa_common.c2000/01/31 20:42:03
@@ -783,8 +783,7 @@
static void
isa_driver_added(device_t dev, driver_t *driver)
{
- device_t *children;
- int nchildren, i;
+ device_t child;
/*
* Don't do anything if drivers are dynamically
@@ -796,11 +795,10 @@
return;
DEVICE_IDENTIFY(driver, dev);
- if (device_get_children(dev, children, nchildren))
- return;
- for (i = 0; i nchildren; i++) {
- device_t child = children[i];
+ for (child = TAILQ_FIRST(dev-children);
+child;
+child = TAILQ_NEXT(child, link)) {
struct isa_device *idev = DEVTOISA(child);
struct resource_list *rl = idev-id_resources;
struct resource_list_entry *rle;
@@ -831,8 +829,6 @@
}
}
}
-
- free(children, M_TEMP);
}
static int
Re: kldloaded driver not called at load time
I am trying to make kldloadable driver. But none of functions are called when I kldload it. Is this a bug or I am missing something essential? You're missing something essential. The return values for the DEVICE_PROBE method should be 0 or negative for priority, or positive to indicate an error. THE PROBLEM is that tok_isa_probe is not called at all, when driver kldloaded. Return value is irrelevant for this case. This is because the unknown driver is never revoked once it owns your device. It's a(nother) bug in the ISA bus code. 8( -- \\ Give a man a fish, and you feed him for a day. \\ Mike Smith \\ Tell him he should learn how to fish himself, \\ [EMAIL PROTECTED] \\ and he'll hate you for a lifetime. \\ [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Mike Smith wrote: This is because the unknown driver is never revoked once it owns your device. It's a(nother) bug in the ISA bus code. 8( Looking at the code in sys/kern/subr_bus.c I'm not sure how any drivers are revoked by a higher priority match. This is probably desirable given the nature of most drivers and one of the reasons for letting BUS_PROBE_NOMATCH announce your 'unknown' hardware. If a driver is attached and providing interfaces to the system (network, scsi, what have you.) then loading a higher priority driver shouldn't cause that device to detach as it would potentially cause the system to become confused. I never liked the unknown driver. I vote we kill it and let isa_probe_nomatch() deal with things. (I'll have to write isa_probe_nomatch() but I don't see that as a problem.) Nikolai, your easy fix is to comment out the relevent module declaration for the 'unknown' driver in isa_common.c. -- | 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: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Matthew N. Dodd wrote: On Mon, 31 Jan 2000, Mike Smith wrote: This is because the unknown driver is never revoked once it owns your device. It's a(nother) bug in the ISA bus code. 8( Looking at the code in sys/kern/subr_bus.c I'm not sure how any drivers are revoked by a higher priority match. This is probably desirable given the nature of most drivers and one of the reasons for letting BUS_PROBE_NOMATCH announce your 'unknown' hardware. If a driver is attached and providing interfaces to the system (network, scsi, what have you.) then loading a higher priority driver shouldn't cause that device to detach as it would potentially cause the system to become confused. I never liked the unknown driver. I vote we kill it and let isa_probe_nomatch() deal with things. (I'll have to write isa_probe_nomatch() but I don't see that as a problem.) Nikolai, your easy fix is to comment out the relevent module declaration for the 'unknown' driver in isa_common.c. This is probably the right workaround for now. The unknown driver does serve a purpose. It indirectly allows resources associated with the device to be allocated so that other PnP devices are not allocated on top. On reviewing the code, this seems to happen (in isa_probe_children) whether or not the preceding call to device_probe_and_attach() succeeded so removing the unknown driver would mean that resources are allocated for the un-attached device which is likely to cause trouble. -- Doug Rabson Mail: [EMAIL PROTECTED] Nonlinear Systems Ltd. Phone: +44 181 442 9037 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
Nikolai, your easy fix is to comment out the relevent module declaration for the 'unknown' driver in isa_common.c. Yes, this fix does the job, BUT Comments for 'unknown' driver says '... for unknown pnp cards ...'. Will my driver be called on computer _without_ any pnp card? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Matthew N. Dodd wrote: On Mon, 31 Jan 2000, Nikolai Saoukh wrote: THE PROBLEM is that tok_isa_probe is not called at all, when driver kldloaded. Return value is irrelevant for this case. Try the attached patch. (cd /sys/isa patch isa_common.c.patch) This patch doesn't (shouldn't) compile since it uses private information from bus_private.h. I hope you are not including bus_private.h anywhere in your code - its strictly off limits to driver code, even bus drivers. -- Doug Rabson Mail: [EMAIL PROTECTED] Nonlinear Systems Ltd. Phone: +44 181 442 9037 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
On Tue, 1 Feb 2000, Nikolai Saoukh wrote: Comments for 'unknown' driver says '... for unknown pnp cards ...'. Will my driver be called on computer _without_ any pnp card? Does your driver have an identify method for non-PnP cards? -- | 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: kldloaded driver not called at load time
On Mon, 31 Jan 2000, Doug Rabson wrote: This patch doesn't (shouldn't) compile since it uses private information from bus_private.h. I hope you are not including bus_private.h anywhere in your code - its strictly off limits to driver code, even bus drivers. Yea, I figured that out. My patch didn't change the function of the code anyway. Not sure what I was thinking. -- | 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: kldloaded driver not called at load time
On Tue, 1 Feb 2000, Nikolai Saoukh wrote: Comments for 'unknown' driver says '... for unknown pnp cards ...'. Will my driver be called on computer _without_ any pnp card? Does your driver have an identify method for non-PnP cards? This particular card can be "downgraded" to previous model (Auto 16/4) in strict legacy mode (no pnp at all). Or with some pnp. Well, my question was not about my case, rather general one. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
"Matthew N. Dodd" wrote: On Mon, 31 Jan 2000, Mike Smith wrote: This is because the unknown driver is never revoked once it owns your device. It's a(nother) bug in the ISA bus code. 8( Looking at the code in sys/kern/subr_bus.c I'm not sure how any drivers are revoked by a higher priority match. This is probably desirable given the nature of most drivers and one of the reasons for letting BUS_PROBE_NOMATCH announce your 'unknown' hardware. I never liked the unknown driver. I vote we kill it and let isa_probe_nomatch() deal with things. (I'll have to write isa_probe_nomatch() but I don't see that as a problem.) I'll dig out my patches for that area and send you a copy. We do need something when we use PNPBIOS because the motherboard devices are not programmable and their port space is "fixed". Their space needs to either be allocated to something or otherwise avoided when we are choosing ports for real programmable devices. Cheers, -Peter To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: kldloaded driver not called at load time
On Tue, 1 Feb 2000, Peter Wemm wrote: We do need something when we use PNPBIOS because the motherboard devices are not programmable and their port space is "fixed". Their space needs to either be allocated to something or otherwise avoided when we are choosing ports for real programmable devices. If we can make the relevent code use RF_ALLOCATED vs. RF_ACTIVE for determining resource availability I think this will work fine. -- | 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
