On 10/04/2010 09:51 PM, Chris Frey wrote:
> On Mon, Oct 04, 2010 at 06:37:36PM +0200, Tejun Heo wrote:
>> Hello, sorry about chiming in later.  I was off last week.
> 
> No problem, I'm eager to test patches to fix this.
> 
>> I think we're on the right track.  The problem with Jens' patch was
>> that it didn't consider the fact that blk_end_request() now internally
>> updates the current position of the request, so if restart happens
>> after some of part of the request is complete it will end up adding
>> the offsets multiple times.  I think slightly modifying it to track
>> the current position instead of offset should do it.  Chris, can you
>> please try the following patch and see whether the problem goes away?
> 
> Unfortunately, this patch does not fix it for me.  I applied your patch
> to kernel 2.6.35.5, and got the usual error:
> 
> EXT3-fs error (device ubda): ext3_lookup: deleted inode referenced: 566188
> 
> Lots of them, actually.

Hmmmm, can you please give a shot at the following one?  Thank you.

diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c
index 1bcd208..0435d21 100644
--- a/arch/um/drivers/ubd_kern.c
+++ b/arch/um/drivers/ubd_kern.c
@@ -163,6 +163,7 @@ struct ubd {
        struct scatterlist sg[MAX_SG];
        struct request *request;
        int start_sg, end_sg;
+       sector_t rq_pos;
 };

 #define DEFAULT_COW { \
@@ -187,6 +188,7 @@ struct ubd {
        .request =              NULL, \
        .start_sg =             0, \
        .end_sg =               0, \
+       .rq_pos =               0, \
 }

 /* Protected by ubd_lock */
@@ -1228,7 +1230,6 @@ static void do_ubd_request(struct request_queue *q)
 {
        struct io_thread_req *io_req;
        struct request *req;
-       sector_t sector;
        int n;

        while(1){
@@ -1239,12 +1240,12 @@ static void do_ubd_request(struct request_queue *q)
                                return;

                        dev->request = req;
+                       dev->rq_pos = blk_rq_pos(req);
                        dev->start_sg = 0;
                        dev->end_sg = blk_rq_map_sg(q, req, dev->sg);
                }

                req = dev->request;
-               sector = blk_rq_pos(req);
                while(dev->start_sg < dev->end_sg){
                        struct scatterlist *sg = &dev->sg[dev->start_sg];

@@ -1256,10 +1257,10 @@ static void do_ubd_request(struct request_queue *q)
                                return;
                        }
                        prepare_request(req, io_req,
-                                       (unsigned long long)sector << 9,
+                                       (unsigned long long)dev->rq_pos << 9,
                                        sg->offset, sg->length, sg_page(sg));

-                       sector += sg->length >> 9;
+                       dev->rq_pos += sg->length >> 9;
                        n = os_write_file(thread_fd, &io_req,
                                          sizeof(struct io_thread_req *));
                        if(n != sizeof(struct io_thread_req *)){

------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
_______________________________________________
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Reply via email to