cvs commit: src/sys/dev/ata atapi-cd.c atapi-cd.h

[Poul-Henning Kamp]

In message <86lk8umd3w.fsf at ds4.des.no>, =?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?= wr
ites:
>"Poul-Henning Kamp" <phk at phk.freebsd.dk> writes:
>> In message <473DD94D.8050605 at samsco.org>, Scott Long writes:
>> > But then why not have g_access be able to communicate the ability to
>> > access the device independent from the ability to access the media?
>> > Seems like both of these questions are "access" questions.
>> Because that is a lot more work for no benefit I can see ?
>
>There is another closely related problem which also needs fixing and
>which also requires GEOM to distinguish between device and media: GEOM's
>behaviour when the media changes while the device is open (e.g. burning
>a CD).  AFAIK, it is not currently possible to properly fixate a CD-R in
>FreeBSD; the CDRIOCFIXATE ioctl always fails because once the media has
>changed (from blank to non-blank), all access fails until you close and
>reopen the device.

This is so CD specific that somehow it must be managed inside the driver.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.