In message <[EMAIL PROTECTED]>, Julian Elischer writes:
>Poul-Henning Kamp wrote:
>> In message <[EMAIL PROTECTED]>, Julian Elischer writes:
>> >
>> >The vmware vmmon driver used to use si_drv1 but suddenly it appears to
>> >have already been used (at least it is no-longer non-zero).
>> >It would fail to open because it checks it before using it during open.
>> >Changing the driver to use si_drv2 fixes the problem.
>> >
>> >Since drv1 is defined as 'for the use of the driver'
>> >I am not sure that it is wrong for the driver to use it.
>> >In fact many other drivers do use it without testing it.
>> >
>> >I see it's used in freedev() and make_dev_alias().
>> >Is it safe for the driver to use it even if it is non-zero?
>> I think you need to give me a bit more details before I an answer
>> meaningfully...
>> Are you sure that the value is not from previous use by vmmon ?
>no I am not sure.. I will test further.
>the sequence:
>1/ failed to open
>2/ unload module

Make sure vmmon calles destroy_dev() on all dev_t's created with
make_dev() or make_dev_alias() when unloaded.

This process is not automatic, and I am not sure we can safely
make it automatic.  We may be able to make it easier though.

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

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Reply via email to