Hi all,
These patches move SIIG serial & parallel combo cards support from
serial and parport drivers to parport_serial.c (where IMHO it must be).
Untested, but compiles and should work :))
These patches were sended to LKML some months ago, but seems like they
was lost somewhere and I can't remember any answer.
Best regards.
--
Andrey Panin | Embedded systems software engineer
[email protected] | PGP key: http://www.orbita1.ru/~pazke/AndreyPanin.asc
On Sat, Dec 15, 2001 at 04:57:39PM +0300, Andrey Panin wrote:
> Untested, but compiles and should work :))
>
> These patches were sended to LKML some months ago, but seems like they
> was lost somewhere and I can't remember any answer.
I'm waiting for someone to tell me that it still works. :-)
Tim.
*/
On Fri, Dec 14, 2001 at 01:57:49PM +0000, Tim Waugh wrote:
> On Sat, Dec 15, 2001 at 04:57:39PM +0300, Andrey Panin wrote:
>
> > Untested, but compiles and should work :))
> >
> > These patches were sended to LKML some months ago, but seems like they
> > was lost somewhere and I can't remember any answer.
>
> I'm waiting for someone to tell me that it still works. :-)
I'm still waiting for people to test out the new serial layer and
report all the bugs that are bound to be in there. 8)
One you're happy with the changes, I'll put them into my serial cvs.
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
I've done some grep'ing for some possible fixes required in 2.5.1 (final).
The problems are:
- changing of member cmd in 'struct request' to a char array from int
- changing of blk_init_queue() to take 3 args not 2
- use of the removed member b_reqnext in 'struct buffer_head'
Note there are probably some files I have missed due to not thinking generally
enough and there will be false matches also. As you can probably tell from the
below I'm definitely not a grep wizard so if someone can think of a way to
search for code like
printk(" ... %d ... ", ..., rq->cmd, ...);
that will turn up more things in need of fixing.
Sorry if this info has already been found - I couldn't find an overall report
like this in the archives.
1)
Look for possible invalid use of altered member cmd from struct request
- search for this kind of thing
switch (rq->cmd)
{
case READ:
...
grep "switch[[:space:]]*([a-zA-Z0-9_]*->cmd[[:space:]]*)"
grep "switch[[:space:]]*([a-zA-Z0-9_]*\.cmd[[:space:]]*)"
then filter on
grep "struct[[:space:]]*request[[:space:]]"
drivers/block/cpqarray.c
drivers/ide/ide-floppy.c
drivers/ide/ide-tape.c
drivers/s390/block/xpram.c
drivers/mtd/mtdblock.c
drivers/mtd/mtdblock_ro.c
----------
2)
Look for possible invalid use of altered member cmd from struct request
- search for this kind of thing
if (rq->cmd == READ) {
...
and
rq->cmd = READ;
grep -- "->cmd[[:space:]]*="
then filter on
grep "struct[[:space:]]*request[[:space:]]"
include/linux/ide.h : Documentation change
drivers/block/nbd.c
drivers/sbus/char/jsflash.c
drivers/ide/icside.c
drivers/ide/ide-cd.c
drivers/ide/ide-floppy.c
drivers/ide/ide-tape.c
drivers/ide/pdc4030.c
drivers/acorn/block/mfmhd.c
drivers/s390/block/dasd.c
drivers/s390/block/dasd_diag.c
drivers/s390/block/dasd_eckd.c
drivers/s390/block/dasd_fba.c
drivers/mtd/nftlcore.c
drivers/message/i2o/i2o_block.c
--------------
3)
Search for use of blk_init_queue() with only two args.
grep -l blk_init_queue[[:space:]]*([^,]*,[^,)]*)
drivers/s390/char/tapedefs.h : OK - compatibility with older kernels.
arch/m68k/atari/stram.c
drivers/block/DAC960.c
drivers/cdrom/aztcd.c
drivers/cdrom/cm206.c
drivers/cdrom/gscd.c
drivers/cdrom/mcd.c
drivers/cdrom/mcdx.c
drivers/cdrom/optcd.c
drivers/cdrom/sbpcd.c
drivers/cdrom/sjcd.c
drivers/cdrom/sonycd535.c
drivers/sbus/char/jsflash.c
drivers/acorn/block/fd1772.c
drivers/acorn/block/mfmhd.c
drivers/s390/block/dasd.c
drivers/s390/block/xpram.c
drivers/s390/char/tapeblock.c
drivers/mtd/ftl.c
drivers/mtd/mtdblock.c
drivers/mtd/mtdblock_ro.c
drivers/mtd/nftlcore.c
drivers/message/i2o/i2o_block.c
-------
4)
Search for use of removed member b_reqnext in struct buffer_head.
fgrep "b_reqnext"
include/linux/raid/raid5.h : documentation change
arch/cris/drivers/ide.c
drivers/block/acsi.c
drivers/scsi/scsi_debug.c
drivers/ide/icside.c
drivers/ide/ide-pmac.c
drivers/ide/ide-tape.c
drivers/s390/block/dasd_diag.c
drivers/s390/block/dasd_eckd.c
drivers/s390/block/dasd_fba.c
drivers/s390/char/tape34xx.c
drivers/s390/char/tapeblock.c
drivers/md/raid5.c
On Fri, Dec 14, 2001 at 04:22:53PM +0000, Russell King wrote:
> On Fri, Dec 14, 2001 at 01:57:49PM +0000, Tim Waugh wrote:
> > On Sat, Dec 15, 2001 at 04:57:39PM +0300, Andrey Panin wrote:
> >
> > > Untested, but compiles and should work :))
> > >
> > > These patches were sended to LKML some months ago, but seems like they
> > > was lost somewhere and I can't remember any answer.
> >
> > I'm waiting for someone to tell me that it still works. :-)
>
> I'm still waiting for people to test out the new serial layer and
> report all the bugs that are bound to be in there. 8)
>
> One you're happy with the changes, I'll put them into my serial cvs.
Can I download it not using cvs, to test it with my PCI serial card and
ISAPNP modem on SMP motherboard ?
--
Andrey Panin | Embedded systems software engineer
[email protected] | PGP key: http://www.orbita1.ru/~pazke/AndreyPanin.asc