Bill Fumerola wrote:
>
> On Wed, Jan 19, 2000 at 03:50:49PM -0700, Wes Peters wrote:
>
> > Mmm... Good point. This means updating code and compiling a new kernel
> > every time Joe's hardware company sticks their own name on a generic OEM
> > PCI or PCCard, doesn't it? I love this industry.
>
On Wed, Jan 19, 2000 at 03:50:49PM -0700, Wes Peters wrote:
> Mmm... Good point. This means updating code and compiling a new kernel
> every time Joe's hardware company sticks their own name on a generic OEM
> PCI or PCCard, doesn't it? I love this industry.
Do the above and try and figure ou
In message <[EMAIL PROTECTED]> Mike Smith writes:
: It can make a pretty good guess though; certainly good enough in most
: cases.
The problem with guesses is that they are guesses and often wrong.
Consider a simple case. If I don't have a sound driver configured on
my laptop, IRQ 5 could appe
> In message <[EMAIL PROTECTED]> "Matthew N.
>Dodd" writes:
> : pccardd shouldn't have to specify a list of IRQs to use; thats info the
> : kernel knows about.
>
> Oops. Missed this part. The problem again is that the kernel doesn't
> know about this. At least it knows it only to a point. I
On 19-Jan-00 Warner Losh wrote:
> In message <[EMAIL PROTECTED]> Wes Peters writes:
> : Do you really want to cram a list of all known PCcard and USB
> : devices
> : into your kernel? Ugh.
> Do you really want to cram a list of all known pci cards into the
> kernel? Same thing really.
So a
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: Too many drivers require knowledge of specific cards in order to properly
: handle 'quirks'. This isn't going to change. While a recompile is
: annoying we can be assured that a driver maintainer will only add IDs to a
: driver after th
On Wed, 19 Jan 2000, Wes Peters wrote:
> Mmm... Good point. This means updating code and compiling a new
> kernel every time Joe's hardware company sticks their own name on a
> generic OEM PCI or PCCard, doesn't it? I love this industry.
Too many drivers require knowledge of specific cards in
Warner Losh wrote:
>
> In message <[EMAIL PROTECTED]> Wes Peters writes:
> : Do you really want to cram a list of all known PCcard and USB devices
> : into your kernel? Ugh.
>
> Do you really want to cram a list of all known pci cards into the
> kernel? Same thing really.
Mmm... Good point.
Warner Losh wrote:
> In message <[EMAIL PROTECTED]> Edwin Culp writes:
> : Thanks, I would not have found it for a while. I checked UPDATING and saw nothing,
> : although my problem may have been because I haven't been able to finish a world
> : since the commit.
>
> That's an oversight on my pa
In message <[EMAIL PROTECTED]> Edwin Culp writes:
: Thanks, I would not have found it for a while. I checked UPDATING and saw nothing,
: although my problem may have been because I haven't been able to finish a world
: since the commit.
That's an oversight on my part. Which I'm correcting right
In message <[EMAIL PROTECTED]> Wes Peters writes:
: Do you really want to cram a list of all known PCcard and USB devices
: into your kernel? Ugh.
Do you really want to cram a list of all known pci cards into the
kernel? Same thing really.
Warner
To Unsubscribe: send mail to [EMAIL PROTECTED
"Matthew N. Dodd" wrote:
>
> Matching drivers to hardware is
> something a newbus driver should do itself; relying on an external hint
> mechanism strikes me as a solution prone to user aggravation. (speaking
> as an aggravated user of course.)
Do you really want to cram a list of all known PCc
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: Ok, I'm obviously missing something here... You mean the IRQs specified
: in /etc/pccard.conf are a complete crapshoot?
Well, yes. The IRQs are supposed to be free AND ROUTABLE TO THE PCIC
part. If they are, then it works great. If
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: Well, I wasn't really trying to flame you, or anyone. It is frusterating
: to see how much time gets tied up in tracking down dodgy PCIC configs
: though. Solving that would probably free you up of several emails a
: week. If I start m
On Wed, 19 Jan 2000, Warner Losh wrote:
> I'd love to find a good solution to this problem. I'm glad your flame
> caught me at a time when my brain was up for trying to solve this
> silly problem.
Well, I wasn't really trying to flame you, or anyone. It is frusterating
to see how much time gets
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: Otherwise your timeout solution seems good. Are the valid IRQs dependent
: on the board integrator or on the chip used? Since we can identify the
: chips (well, it looks like we can) couldn't we maintain a 'quirk' table?
That I'm not s
On Wed, 19 Jan 2000, Warner Losh wrote:
> Well, yes. The IRQs are supposed to be free AND ROUTABLE TO THE PCIC
> part. If they are, then it works great. If they aren't, then we lose
> bigtime.
It seems to me that any IRQ that the PCIC can see will be usable for
assignment to a card no? Is the
On Tue, 18 Jan 2000, Warner Losh wrote:
> Oops. Missed this part. The problem again is that the kernel doesn't
> know about this. At least it knows it only to a point. It knows
> which IRQs are in use, but it doesn't know if the pcic (or cardbus
> bridge in compat mode) can route to a given fr
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: pccardd shouldn't have to specify a list of IRQs to use; thats info the
: kernel knows about.
Oops. Missed this part. The problem again is that the kernel doesn't
know about this. At least it knows it only to a point. It knows
which
In message <[EMAIL PROTECTED]> "Matthew N. Dodd"
writes:
: The pcic driver should be smart enough to find a free IRQ; in addition
: pccardd shouldn't have to specify a list of IRQs to use; thats info the
: kernel knows about.
Should be smart enough? How? There are a limited number of IRQs
avai
On Tue, 18 Jan 2000, Warner Losh wrote:
> Also, you will likely need to add an IRQ to the pcic line in your
> kernel config.
The pcic driver should be smart enough to find a free IRQ; in addition
pccardd shouldn't have to specify a list of IRQs to use; thats info the
kernel knows about.
I'm ki
In message <[EMAIL PROTECTED]> Edwin Culp writes:
: :Jan 17 23:33:58 local-27 pccardd[46]: driver allocation failed for
: D-Link(DE-66
: 0): Inappropriate ioctl for device
Make sure that your kernel and pccardd match. The binary api changes
from time to time in -curent. This is a sure symptom.
Thank you very much. The include file was different and I copied it and
rebuilt pccardc and pccardd and think that will solve the problem.
Thanks,
ed
MIHIRA Yoshiro wrote:
> [EMAIL PROTECTED] wrote:
>
> >> I am running current as of yesterday. I have neglegently not been
> >> rebooting after
ld not have found it for a while. I checked UPDATING and saw nothing,
although my problem may have been because I haven't been able to finish a world
since the commit.
ed
Masahide -mac- NODA wrote:
> From: Edwin Culp <[EMAIL PROTECTED]>
> Subject: Re: Problems with PCMCIA Ca
[EMAIL PROTECTED] wrote:
>> I am running current as of yesterday. I have neglegently not been
>> rebooting after make world and making a new kernel. Yesterday I did
>> reboot and found that something has changed and my D-Link 660 and my
>> Viking Modem PCMCIA cards are no longer working.
>>
>>
From: Edwin Culp <[EMAIL PROTECTED]>
Subject: Re: Problems with PCMCIA Cards
Date: Tue, 18 Jan 2000 16:45:25 -0800
Message-ID: <[EMAIL PROTECTED]>
eculp> > On Tue, Jan 18, 2000 at 11:31:17AM -0800, Edwin Culp wrote:
eculp> > > I am running current as of yesterday.
Oscar,
Thanks a lot, but my kernel.conf has
device pcic0 isa?
Already.
Thanks,
ed
Oscar Bonilla wrote:
> On Tue, Jan 18, 2000 at 11:31:17AM -0800, Edwin Culp wrote:
> > I am running current as of yesterday. I have neglegently not been
> > rebooting after make world and making a new kernel.
On Tue, Jan 18, 2000 at 11:31:17AM -0800, Edwin Culp wrote:
> I am running current as of yesterday. I have neglegently not been
> rebooting after make world and making a new kernel. Yesterday I did
> reboot and found that something has changed and my D-Link 660 and my
> Viking Modem PCMCIA cards
I am running current as of yesterday. I have neglegently not been
rebooting after make world and making a new kernel. Yesterday I did
reboot and found that something has changed and my D-Link 660 and my
Viking Modem PCMCIA cards are no longer working.
:Jan 17 23:33:58 local-27 pccardd[46]: driv
29 matches
Mail list logo
