Hiya.
The latest reiserfs patch on ftp.namesys.com causes compilation errors
against test12 due to the task queue changes. Does this look correct?
--- fs/reiserfs/journal.c.orig Wed Dec 13 00:13:00 2000
+++ fs/reiserfs/journal.c Wed Dec 13 00:40:52 2000
@@ -1762,7 +1762,7 @@
ct->p_s_sb = p_s_sb ;
ct->jindex = jindex ;
ct->task_done = NULL ;
- ct->task.next = NULL ;
+ INIT_LIST_HEAD(&ct->task.list);
ct->task.sync = 0 ;
ct->task.routine = (void *)(void *)reiserfs_journal_commit_task_func ;
ct->self = ct ;
@@ -1813,7 +1813,7 @@
lock_kernel() ;
while(1) {
- while(reiserfs_commit_thread_tq) {
+ while(TQ_ACTIVE(reiserfs_commit_thread_tq)) {
run_task_queue(&reiserfs_commit_thread_tq) ;
}
It does at least compile with these changes, but I haven't yet tested
it. Looking at run_task_queue, it would appear that the while() is now
redundant, but could someone who knows confirm/deny this?
--
Adam Sampson
[email protected]
Adam Sampson wrote:
>The latest reiserfs patch on ftp.namesys.com causes compilation errors
>against test12 due to the task queue changes. Does this look correct?
[patch snipped]
>
>It does at least compile with these changes, but I haven't yet tested
>it. Looking at run_task_queue, it would appear that the while() is now
>redundant, but could someone who knows confirm/deny this?
Chris Mason is working on this. In an earlier exchange on reiserfs-list:
Chris Mason wrote:
>I'll try to hack out a patch while I'm waiting, but the task struct changes
>are the least of our problems. ?They've changed ll_rw_block to always set
>the end_io handler to the default one. ?Since the journal code relies on
>being able to use its own end_io handler, this isn't good for us. ?There is
>a new func we need to use instead, so I'm migrating over.
>
>Thew new stuff should be faster, so I won't complain ;-)
Your patch will probably let journal.c get compiled, but it might be dangerous
to use considering what Chris said.
Steven
Steven Cole <[email protected]> writes:
> Your patch will probably let journal.c get compiled, but it might be
> dangerous to use considering what Chris said.
I've just tried it and got filesystem corruption, so no, it doesn't
work. I'll wait for someone else to provide a patch that works.
--
Adam Sampson
[email protected]
> Hiya.
>
> The latest reiserfs patch on ftp.namesys.com causes compilation errors
> against test12 due to the task queue changes. Does this look correct?
>
Please, somebody can test this patch
-- BOF --
diff -ur linux-bk/fs/reiserfs/journal.c linux/fs/reiserfs/journal.c
--- linux-bk/fs/reiserfs/journal.c Tue Dec 12 14:38:22 2000
+++ linux/fs/reiserfs/journal.c Tue Dec 12 17:41:19 2000
@@ -77,7 +77,9 @@
/* wait on this if you need to be sure you task queue entries have been run */
static DECLARE_WAIT_QUEUE_HEAD(reiserfs_commit_thread_done) ;
-DECLARE_TASK_QUEUE(reiserfs_commit_thread_tq) ;
+
+/*DECLARE_TASK_QUEUE(reiserfs_commit_thread_tq) ; */
+task_queue *reiserfs_commit_thread_tq = NULL;
#define JOURNAL_TRANS_HALF 1018 /* must be correct to keep the desc and commit structs at 4k */
@@ -1762,7 +1764,7 @@
ct->p_s_sb = p_s_sb ;
ct->jindex = jindex ;
ct->task_done = NULL ;
- ct->task.next = NULL ;
+ INIT_LIST_HEAD(&ct->task.list);
ct->task.sync = 0 ;
ct->task.routine = (void *)(void *)reiserfs_journal_commit_task_func ;
ct->self = ct ;
@@ -1777,7 +1779,7 @@
ct = kmalloc(sizeof(struct reiserfs_journal_commit_task), GFP_BUFFER) ;
if (ct) {
setup_commit_task_arg(ct, p_s_sb, jindex) ;
- queue_task(&(ct->task), &reiserfs_commit_thread_tq);
+ queue_task(&(ct->task), reiserfs_commit_thread_tq);
wake_up(&reiserfs_commit_thread_wait) ;
} else {
#ifdef CONFIG_REISERFS_CHECK
@@ -1813,13 +1814,13 @@
lock_kernel() ;
while(1) {
- while(reiserfs_commit_thread_tq) {
- run_task_queue(&reiserfs_commit_thread_tq) ;
+ while(reiserfs_commit_thread_tq) {
+ run_task_queue(reiserfs_commit_thread_tq) ;
}
/* if there aren't any more filesystems left, break */
if (reiserfs_mounted_fs_count <= 0) {
- run_task_queue(&reiserfs_commit_thread_tq) ;
+ run_task_queue(reiserfs_commit_thread_tq) ;
break ;
}
wake_up(&reiserfs_commit_thread_done) ;
diff -ur linux-bk/include/linux/reiserfs_fs.h linux/include/linux/reiserfs_fs.h
--- linux-bk/include/linux/reiserfs_fs.h Tue Dec 12 14:39:34 2000
+++ linux/include/linux/reiserfs_fs.h Tue Dec 12 17:31:09 2000
@@ -1565,8 +1565,8 @@
int do_not_lock ;
} ;
-extern task_queue reiserfs_commit_thread_tq ;
-extern task_queue reiserfs_end_io_tq ;
+extern task_queue *reiserfs_commit_thread_tq ;
+extern task_queue *reiserfs_end_io_tq ;
extern wait_queue_head_t reiserfs_commit_thread_wait ;
/* biggest tunable defines are right here */
-- EOF --
cya;
--
_________________________
Carlos E Gorges
([email protected])
Tech inform?tica LTDA
Brazil
_________________________
Hi Carlos!
I have this oops with your patch. applied to clean 2.4.0test12+reiserfs
3.6.22
reiserfs is compiled as module. I can mount the partition but after 4 o
5 seconds kupdate oops, dies and I have to reboot becauser I can no longer
write to any disk. The system didn't crashed I can open new terminal and run
programs.
Please don't hesitate to contact me if you need more info.
-Jorge
---------------------------------------------------------------------
ksymoops 2.3.4 on i686 2.4.0-test12. Options used
-V (specified)
-k /proc/ksyms (specified)
-l /proc/modules (specified)
-o /lib/modules/2.4.0-test12/ (specified)
-m /boot/System.map-2.4.0-test12 (specified)
Unable to handle kernel NULL pointer dereference at virtual address 00000004
d084d180
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[reiserfs:__insmod_reiserfs_S.text_L121821+106784/121888]
EFLAGS: 00010046
eax: 00000246 ebx: cf734680 ecx: 00000000 edx: cf734000
esi: cf734694 edi: 00000000 ebp: 00000000 esp: c15e7f0c
ds: 0018 es: 0018 ss: 0018
Process kupdate (pid: 6, stackpage=c15e7000)
Stack: 00000000 00000012 00000010 d084f450 c1550e00 00000000 c15e7f98
c15e623a
d0856916 c1550e00 00003ce1 00000000 00000001 00000000 0000000c
0000000b
00000000 00001cd6 cf9964c0 cfe57340 ceaed000 cf8de000 d08782f4
d084e21e
Call Trace: [reiserfs:__insmod_reiserfs_S.text_L121821+115696/121888]
[reiserfs:__insmod_reiserfs_S.rodata_L26959+23702/26976]
[scsi_mod:proc_scsi_Rba5e7901+345560/74510060]
[reiserfs:__insmod_reiserfs_S.text_L121821+111038/121888]
[reiserfs:__insmod_reiserfs_S.rodata_L26959+23279/26976]
[reiserfs:__insmod_reiserfs_S.text_L121821+50184/121888] [tvecs+13774/62872]
Code: 8b 51 04 89 71 04 89 4b 14 89 56 04 89 32 50 9d b9 01 00 00
Using defaults from ksymoops -t elf32-i386 -a i386
Code; 00000000 Before first symbol
00000000 <_EIP>:
Code; 00000000 Before first symbol
0: 8b 51 04 mov 0x4(%ecx),%edx
Code; 00000003 Before first symbol
3: 89 71 04 mov %esi,0x4(%ecx)
Code; 00000006 Before first symbol
6: 89 4b 14 mov %ecx,0x14(%ebx)
Code; 00000009 Before first symbol
9: 89 56 04 mov %edx,0x4(%esi)
Code; 0000000c Before first symbol
c: 89 32 mov %esi,(%edx)
Code; 0000000e Before first symbol
e: 50 push %eax
Code; 0000000f Before first symbol
f: 9d popf
Code; 00000010 Before first symbol
10: b9 01 00 00 00 mov $0x1,%ecx
----------------------------------------------------------------
==============================================================
Jorge Boncompte - T?cnico de sistemas
DTI2 - Desarrollo de la Tecnolog?a de las Comunicaciones
--------------------------------------------------------------
C/ Abogado Enriquez Barrios, 5 14004 CORDOBA (SPAIN)
Tlf: +34 957 761395 / FAX: +34 957 450380
--------------------------------------------------------------
[email protected] _-_-_-_-_-_-_-_-_-_-_-_-_-_ http://www.dti2.net
==============================================================
Without wicker a basket cannot be done.
==============================================================
----- Mensaje original -----
De: "Carlos E. Gorges" <[email protected]>
Para: <[email protected]>
Enviado: mi?rcoles, 13 de diciembre de 2000 14:14
Asunto: Re: Possible patch for reiserfs-3.6.22 against 2.4.0-test12
> > Hiya.
> >
> > The latest reiserfs patch on ftp.namesys.com causes compilation errors
> > against test12 due to the task queue changes. Does this look correct?
> >
>
> Please, somebody can test this patch
>
> -- BOF --
> diff -ur linux-bk/fs/reiserfs/journal.c linux/fs/reiserfs/journal.c
> --- linux-bk/fs/reiserfs/journal.c Tue Dec 12 14:38:22 2000
> +++ linux/fs/reiserfs/journal.c Tue Dec 12 17:41:19 2000
> @@ -77,7 +77,9 @@
>
> /* wait on this if you need to be sure you task queue entries have been
run */
> static DECLARE_WAIT_QUEUE_HEAD(reiserfs_commit_thread_done) ;
> -DECLARE_TASK_QUEUE(reiserfs_commit_thread_tq) ;
> +
> +/*DECLARE_TASK_QUEUE(reiserfs_commit_thread_tq) ; */
> +task_queue *reiserfs_commit_thread_tq = NULL;
>
> #define JOURNAL_TRANS_HALF 1018 /* must be correct to keep the desc and
commit structs at 4k */
>
> @@ -1762,7 +1764,7 @@
> ct->p_s_sb = p_s_sb ;
> ct->jindex = jindex ;
> ct->task_done = NULL ;
> - ct->task.next = NULL ;
> + INIT_LIST_HEAD(&ct->task.list);
> ct->task.sync = 0 ;
> ct->task.routine = (void *)(void *)reiserfs_journal_commit_task_func ;
> ct->self = ct ;
> @@ -1777,7 +1779,7 @@
> ct = kmalloc(sizeof(struct reiserfs_journal_commit_task), GFP_BUFFER) ;
> if (ct) {
> setup_commit_task_arg(ct, p_s_sb, jindex) ;
> - queue_task(&(ct->task), &reiserfs_commit_thread_tq);
> + queue_task(&(ct->task), reiserfs_commit_thread_tq);
> wake_up(&reiserfs_commit_thread_wait) ;
> } else {
> #ifdef CONFIG_REISERFS_CHECK
> @@ -1813,13 +1814,13 @@
> lock_kernel() ;
> while(1) {
>
> - while(reiserfs_commit_thread_tq) {
> - run_task_queue(&reiserfs_commit_thread_tq) ;
> + while(reiserfs_commit_thread_tq) {
> + run_task_queue(reiserfs_commit_thread_tq) ;
> }
>
> /* if there aren't any more filesystems left, break */
> if (reiserfs_mounted_fs_count <= 0) {
> - run_task_queue(&reiserfs_commit_thread_tq) ;
> + run_task_queue(reiserfs_commit_thread_tq) ;
> break ;
> }
> wake_up(&reiserfs_commit_thread_done) ;
> diff -ur linux-bk/include/linux/reiserfs_fs.h
linux/include/linux/reiserfs_fs.h
> --- linux-bk/include/linux/reiserfs_fs.h Tue Dec 12 14:39:34 2000
> +++ linux/include/linux/reiserfs_fs.h Tue Dec 12 17:31:09 2000
> @@ -1565,8 +1565,8 @@
> int do_not_lock ;
> } ;
>
> -extern task_queue reiserfs_commit_thread_tq ;
> -extern task_queue reiserfs_end_io_tq ;
> +extern task_queue *reiserfs_commit_thread_tq ;
> +extern task_queue *reiserfs_end_io_tq ;
> extern wait_queue_head_t reiserfs_commit_thread_wait ;
>
> /* biggest tunable defines are right here */
> -- EOF --
>
> cya;
>
> --
> _________________________
> Carlos E Gorges
> ([email protected])
> Tech inform?tica LTDA
> Brazil
> _________________________
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> Please read the FAQ at http://www.tux.org/lkml/
>