[lkml]   [2003]   [Feb]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Patch in this message
SubjectRE: [PATCH] PC-9800 subarch. support for 2.5.62-AC1 (16/21) SCSI
-----Original Message-----
From: Christoph Hellwig
To: Osamu Tomita
Cc: Linux Kernel Mailing List; Alan Cox; Christoph Hellwig
Sent: 2003/02/23 19:52
Subject: Re: [PATCH] PC-9800 subarch. support for 2.5.62-AC1 (16/21) SCSI

>> +int pc98_bios_param(struct block_device *bdev, int *ip)
>> +{
>> + /* Note: This function is called from fs/partitions/nec98.c too. */
>> + /* So we creat 'sdp' from 'bdev' here. */
>> + struct scsi_disk *sdkp = scsi_disk(bdev->bd_disk);
> this is still not good - you shouldn't expose struct scsi_disk outside
> sd.c. Please change the pc98_bios_param() prototype to that of the
> bios_param entry point (direct passing of capacity).
This is solved by using ioctl_by_bdev() in your suggestion. Thanks.

> Can you explain what this first_real_host() stuff is for - we need some
> way to handle this better.
PC98 BIOS create geometry table on boottime orderd by SCSI ID.
We read that to get geometry. If ide-scsi exist we mis-read table.

(Snipped, but I'll fix them. Thanks.)

>> + BIOS_PARAM_OVERRIDE(sdp, bdev, sdkp->capacity, diskinfo);
>> +
> the way this is done is ugly. I'm still not sure how this is done
> best. When do you need the pc98 geometry exactly? i.e. can it happen
> with one of the existing linux scsi drivers?
We need BIOS geometry exactly to create a partion on linux (by fdisk or
GNU/parted). BIOS uses C/H/S access on boottime according to partition
table. If BIOS geometry is not exact, fail to boot from the partition.
I recived the report about this problem from people using advansys driver
without PC98 patch.

>> +#if defined(CONFIG_SCSI_PC980155) ||
>> +#include "pc980155regs.h"
>> +#else /* !CONFIG_SCSI_PC980155 */
>> static inline uchar read_wd33c93(const wd33c93_regs regs, uchar
>> {
>> @@ -203,6 +206,7 @@
>> *regs.SCMD = cmd;
>> mb();
>> }
>> +#endif /* CONFIG_SCSI_PC980155 */
> The wd33c93 changes are ugly as hell, but that's not your fault. I'll
> try to rework it to abstract out the different implementations better.
> Could you perform some testing for me if I send you updated versions?
Yes, of course.

Osamu Tomita
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:33    [W:0.062 / U:3.644 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site