Hi,
This is a patch to fix releasing memory and io regions for the
aic7xxx driver. This applies both 2.4- and 2.5-series.
Without this, you will fail to hot-remove the device.
This patch is tested on an IA32 server with ACPI PCI hotplug,
and reported to work.
--- aic7xxx_linux_pci.c.orig Tue Nov 13 09:19:41 2001
+++ aic7xxx_linux_pci.c Wed Jul 17 18:03:51 2002
@@ -98,6 +98,10 @@
break;
}
}
+#ifdef MMAPIO
+ release_mem_region(pci_resource_start(pdev, 1), 0x1000);
+#endif
+ release_region(pci_resource_start(pdev, 0), 256);
}
#endif /* !LINUX_VERSION_CODE < KERNEL_VERSION(2,4,0) */
Thanks,
--
KOCHI, Takayoshi <[email protected]/[email protected]>
>Hi,
>
>This is a patch to fix releasing memory and io regions for the
>aic7xxx driver. This applies both 2.4- and 2.5-series.
I don't recall when exactly this was fixed in the aic7xxx driver,
but probably 6.2.5 or so. The 2.5.X kernel must not be using
a recent version of the driver. Marcelo's tree has 6.2.8
which definitely does not require this patch (won't even apply).
--
Justin
On Wed, 24 Jul 2002 16:28:48 -0600
"Justin T. Gibbs" <[email protected]> wrote:
> >Hi,
> >
> >This is a patch to fix releasing memory and io regions for the
> >aic7xxx driver. This applies both 2.4- and 2.5-series.
>
> I don't recall when exactly this was fixed in the aic7xxx driver,
> but probably 6.2.5 or so. The 2.5.X kernel must not be using
> a recent version of the driver. Marcelo's tree has 6.2.8
> which definitely does not require this patch (won't even apply).
Oops, I should have to check the latest 2.4.19-rc series first.
I assumed that 2.5.x contains the latest.
Sorry for bothering you and thanks.
Thanks,
--
KOCHI, Takayoshi <[email protected]/[email protected]>
On Wed, Jul 24 2002, Justin T. Gibbs wrote:
> >Hi,
> >
> >This is a patch to fix releasing memory and io regions for the
> >aic7xxx driver. This applies both 2.4- and 2.5-series.
>
> I don't recall when exactly this was fixed in the aic7xxx driver,
> but probably 6.2.5 or so. The 2.5.X kernel must not be using
> a recent version of the driver. Marcelo's tree has 6.2.8
You make it sounds as if someone would be updating it for you. The
version that is in 2.5 is the version that you last updated it to, end
of story.
I've offered to update the 2.5 version before, without much luck. I'll
do the same again.
--
Jens Axboe
>> I don't recall when exactly this was fixed in the aic7xxx driver,
>> but probably 6.2.5 or so. The 2.5.X kernel must not be using
>> a recent version of the driver. Marcelo's tree has 6.2.8
>
>You make it sounds as if someone would be updating it for you. The
>version that is in 2.5 is the version that you last updated it to, end
>of story.
You make it sound like I have ever done any developement for 2.5. I
haven't. Someone else did the port of the aic7xxx to 2.5. End of
story. 8-)
Unfortunately, I haven't had any spare time to play with 2.5. I have
faithfully maintained the 2.4 driver and will look at 2.5 once the
time to do so presents itself.
--
Justin
On Thu, Jul 25 2002, Justin T. Gibbs wrote:
> >> I don't recall when exactly this was fixed in the aic7xxx driver,
> >> but probably 6.2.5 or so. The 2.5.X kernel must not be using
> >> a recent version of the driver. Marcelo's tree has 6.2.8
> >
> >You make it sounds as if someone would be updating it for you. The
> >version that is in 2.5 is the version that you last updated it to, end
> >of story.
>
> You make it sound like I have ever done any developement for 2.5. I
> haven't. Someone else did the port of the aic7xxx to 2.5. End of
> story. 8-)
Heh yes I know, oh and I did the 2.5 "port" (it wasn't much of a port at
that point, but aic7xxx was one of my bio test victims) :-)
> Unfortunately, I haven't had any spare time to play with 2.5. I have
> faithfully maintained the 2.4 driver and will look at 2.5 once the
> time to do so presents itself.
So you have no problem with me updating 2.5 aic7xxx to match 2.4
current?
--
Jens Axboe
>> Unfortunately, I haven't had any spare time to play with 2.5. I have
>> faithfully maintained the 2.4 driver and will look at 2.5 once the
>> time to do so presents itself.
>
>So you have no problem with me updating 2.5 aic7xxx to match 2.4
>current?
Did you ask when you did the first port? 8-)
In otherwords... be my guest.
--
Justin
On Thu, Jul 25 2002, Justin T. Gibbs wrote:
> >> Unfortunately, I haven't had any spare time to play with 2.5. I have
> >> faithfully maintained the 2.4 driver and will look at 2.5 once the
> >> time to do so presents itself.
> >
> >So you have no problem with me updating 2.5 aic7xxx to match 2.4
> >current?
>
> Did you ask when you did the first port? 8-)
In all fairness, the first port _had_ to be done from my point of view
since I needed _something_ to test the changes on. And to defend myself
even further, I didn't have time to ask maintainers permission before
Linus pulled the changes in.
I can probably come up with a handful more reasons if needed :-)
> In otherwords... be my guest.
Cool, thanks.
--
Jens Axboe
>> Did you ask when you did the first port? 8-)
>
>In all fairness, the first port _had_ to be done from my point of view
>since I needed _something_ to test the changes on. And to defend myself
>even further, I didn't have time to ask maintainers permission before
>Linus pulled the changes in.
>
>I can probably come up with a handful more reasons if needed :-)
This is opensource. Once the code goes out, I have limited control
over what people do with it. This is by design and expected. There's
no need to defend yourself since you didn't do anything wrong.
--
Justin
On Thu, Jul 25 2002, Justin T. Gibbs wrote:
> >> Did you ask when you did the first port? 8-)
> >
> >In all fairness, the first port _had_ to be done from my point of view
> >since I needed _something_ to test the changes on. And to defend myself
> >even further, I didn't have time to ask maintainers permission before
> >Linus pulled the changes in.
> >
> >I can probably come up with a handful more reasons if needed :-)
>
> This is opensource. Once the code goes out, I have limited control
> over what people do with it. This is by design and expected. There's
> no need to defend yourself since you didn't do anything wrong.
Well in theory you are right. But I always like to pass changes on to
the maintainer for submission, and I expect others to do likewise. The
maintainer usually has the better grasp of the code and can make the
better call on what to include, reject, etc. Even though it's open
source it doesn't have to be anarchy.
--
Jens Axboe