[PATCH] remove use of get_gendisk in blkfront

[PATCH] remove use of get_gendisk in blkfront

[Xiaofeng Ling]

[-- Attachment #1: Type: text/plain, Size: 283 bytes --]

remove use of get_gendisk in blkfront
gen_gendisk is no long exported by 2.6 kernel. this patch remove the 
call of get_gendisk by saving the gd point in blkfront_info in vbd.c so 
that blkfront can be built as kerenl module.

Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>


[-- Attachment #2: get_gendisk.patch --]
[-- Type: text/x-patch, Size: 1485 bytes --]

diff -r 864434653dab linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c	
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c
@@ -660,6 +651,7 @@
 	info->vdevice = vdevice;
 	info->connected = BLKIF_STATE_DISCONNECTED;
 	info->mi = NULL;
+ 	info->gd = NULL;
 	INIT_WORK(&info->work, blkif_restart_queue, (void *)info);
 
 	info->shadow_free = 0;
diff -r 864434653dab linux-2.6-xen-sparse/drivers/xen/blkfront/block.h
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/block.h	
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/block.h
@@ -115,6 +115,7 @@
 	/* We watch the backend */
 	struct xenbus_watch watch;
 	dev_t dev;
+ 	struct gendisk *gd;
 	int vdevice;
 	blkif_vdev_t handle;
 	int connected;
diff -r 864434653dab linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c
@@ -266,6 +257,7 @@
 		gd->flags |= GENHD_FL_CD;
 
 	add_disk(gd);
+	info->gd = gd;
 
 	return 0;
 
@@ -298,23 +290,19 @@
 void
 xlvbd_del(struct blkfront_info *info)
 {
-	struct block_device *bd;
 	struct gendisk *gd;
 	int unused;
 	request_queue_t *rq;
 
-	bd = bdget(info->dev);
-	if (bd == NULL)
+	gd = info->gd;
+	if (gd == NULL)
 		return;
 
-	gd = get_gendisk(info->dev, &unused);
 	rq = gd->queue;
-
 	del_gendisk(gd);
 	put_disk(gd);
 	xlbd_put_major_info(info->mi);
 	info->mi = NULL;
 	blk_cleanup_queue(rq);
 
-	bdput(bd);
-}
+}

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: [PATCH] remove use of get_gendisk in blkfront

[Jeremy Katz]

[-- Attachment #1: Type: text/plain, Size: 610 bytes --]

On Fri, 2005-09-02 at 16:28 +0800, Xiaofeng Ling wrote:
> remove use of get_gendisk in blkfront
> gen_gendisk is no long exported by 2.6 kernel. this patch remove the 
> call of get_gendisk by saving the gd point in blkfront_info in vbd.c so 
> that blkfront can be built as kerenl module.
> 
> Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>

This looks correct to me and helps in building blkfront as a module.
Any chance of getting it committed?  I've gone ahead and rediffed
against the current tip and cleaned up the warnings.  Attached here.

Signed-off-by: Jeremy Katz <katzj@redhat.com>

Jeremy

[-- Attachment #2: xen-get-gendisk.patch --]
[-- Type: text/x-patch, Size: 1704 bytes --]

diff -r abbe3df33774 linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c	Tue Nov  8 17:39:58 2005
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/blkfront.c	Tue Nov  8 22:49:53 2005
@@ -641,6 +641,7 @@
 	info->vdevice = vdevice;
 	info->connected = BLKIF_STATE_DISCONNECTED;
 	info->mi = NULL;
+ 	info->gd = NULL;
 	INIT_WORK(&info->work, blkif_restart_queue, (void *)info);
 
 	info->shadow_free = 0;
diff -r abbe3df33774 linux-2.6-xen-sparse/drivers/xen/blkfront/block.h
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/block.h	Tue Nov  8 17:39:58 2005
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/block.h	Tue Nov  8 22:49:53 2005
@@ -116,6 +116,7 @@
 	/* We watch the backend */
 	struct xenbus_watch watch;
 	dev_t dev;
+ 	struct gendisk *gd;
 	int vdevice;
 	blkif_vdev_t handle;
 	int connected;
diff -r abbe3df33774 linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c
--- a/linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c	Tue Nov  8 17:39:58 2005
+++ b/linux-2.6-xen-sparse/drivers/xen/blkfront/vbd.c	Tue Nov  8 22:49:53 2005
@@ -259,6 +259,7 @@
 		gd->flags |= GENHD_FL_CD;
 
 	add_disk(gd);
+	info->gd = gd;
 
 	return 0;
 
@@ -292,22 +293,19 @@
 xlvbd_del(struct blkfront_info *info)
 {
 	struct block_device *bd;
-	struct gendisk *gd;
-	int unused;
-	request_queue_t *rq;
 
 	bd = bdget(info->dev);
 	if (bd == NULL)
 		return;
 
-	gd = get_gendisk(info->dev, &unused);
-	rq = gd->queue;
-
-	del_gendisk(gd);
-	put_disk(gd);
+	if (info->gd == NULL)
+		return;
+
+	del_gendisk(info->gd);
+	put_disk(info->gd);
 	xlbd_put_major_info(info->mi);
 	info->mi = NULL;
-	blk_cleanup_queue(rq);
+	blk_cleanup_queue(info->rq);
 
 	bdput(bd);
 }

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: [PATCH] remove use of get_gendisk in blkfront

[Ian Pratt]

> On Fri, 2005-09-02 at 16:28 +0800, Xiaofeng Ling wrote:
> > remove use of get_gendisk in blkfront
> > gen_gendisk is no long exported by 2.6 kernel. this patch 
> remove the 
> > call of get_gendisk by saving the gd point in blkfront_info 
> in vbd.c 
> > so that blkfront can be built as kerenl module.
> > 
> > Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
> 
> This looks correct to me and helps in building blkfront as a module.
> Any chance of getting it committed?  I've gone ahead and 
> rediffed against the current tip and cleaned up the warnings. 
>  Attached here.

Looks good to me -- AFAIK it was only dropped by accident first time
around.

Folks: if you submit a patch for inclusion please keep resending every
48h until you get some kind of response or it's committed.

[Aside: it would be really nice if there was some tool that pulled
patches off a mailing list and fed them into something like bugzilla,
along with all the coments associated with a patch. It should also watch
the changelog list and update the status as 'APPLIED' when appropriate.
At any time an author could withdraw/supercede a patch, and a maintainer
could reject/keep-on-file. A really smart system would also keep patches
fresh by trying to rebase them, or at least note if they still apply
cleanly.]

Thanks,
Ian 

Re: [PATCH] remove use of get_gendisk in blkfront

[Josh Boyer]

On 11/9/05, Ian Pratt <m+Ian.Pratt@cl.cam.ac.uk> wrote:
>
> [Aside: it would be really nice if there was some tool that pulled
> patches off a mailing list and fed them into something like bugzilla,
> along with all the coments associated with a patch. It should also watch
> the changelog list and update the status as 'APPLIED' when appropriate.
> At any time an author could withdraw/supercede a patch, and a maintainer
> could reject/keep-on-file. A really smart system would also keep patches
> fresh by trying to rebase them, or at least note if they still apply
> cleanly.]

Jeremy Kerr wrote a tool called patchwork that sort of does that. 
It's not nearly as full featured as you'd like, but the basic idea is
there.

The project page is at:
http://www.ozlabs.org/~jk/projects/patchwork/

And an example of it running is at:
http://patchwork.ozlabs.org/linuxppc64/

josh

RE: [PATCH] remove use of get_gendisk in blkfront

[Ling, Xiaofeng]

Ian, this is one of the patches to enable para-driver in unmodified linux.
how about the other patches now?
Any comments for them?

Ian Pratt <> wrote:
>> On Fri, 2005-09-02 at 16:28 +0800, Xiaofeng Ling wrote:
>>> remove use of get_gendisk in blkfront gen_gendisk is no long
>>> exported by 2.6 kernel. this patch
>> remove the
>>> call of get_gendisk by saving the gd point in blkfront_info in vbd.c
>>> so that blkfront can be built as kerenl module.
>>> 
>>> Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
>> 
>> This looks correct to me and helps in building blkfront as a module.
>> Any chance of getting it committed?  I've gone ahead and rediffed
>> against the current tip and cleaned up the warnings.
>>  Attached here.
> 
> Looks good to me -- AFAIK it was only dropped by accident first time
> around. 
> 
> Folks: if you submit a patch for inclusion please keep resending
> every 48h until you get some kind of response or it's committed. 
> 
> [Aside: it would be really nice if there was some tool that pulled
> patches off a mailing list and fed them into something like bugzilla,
> along with all the coments associated with a patch. It should also
> watch the changelog list and update the status as 'APPLIED' when
> appropriate. At any time an author could withdraw/supercede a patch,
> and a maintainer could reject/keep-on-file. A really smart system
> would also keep patches fresh by trying to rebase them, or at least
> note if they still apply cleanly.]      
> 
> Thanks,
> Ian
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel