cvs commit: src/sys/kern kern_shutdown.c vfs_subr.c

Nate Lawson nate at
Thu Jul 15 18:07:30 PDT 2004

Alfred Perlstein wrote:
> * John Baldwin <jhb at> [040715 11:56] wrote:
>>On Thursday 15 July 2004 01:29 am, Alfred Perlstein wrote:
>>>>  Also, why do you copy the proc name to the stack lower in
>>>>  kern_shutdown.c?  Do you fear that the proc might change from
>>>>  under you when you're at the bottom and you can't therefore
>>>>  only keep a reference to the proc instead?
>>>Because I don't know if the proc will be there after it nukes itself.
>>They don't nuke themselves, they just go to sleep.  A quick look at the code 
>>would have revealed this.
> y'know what, in the context of fixing a printf I wasn't really
> interested in looking that far into the scheduler.  A 20 char stack
> variable and a string copy at shutdown isn't such a big deal.  If it
> bugs you guys that much, why don't you just fix it?
> Or are you worried about getting 15 emails about how you inefficiently
> printing a non-critical string like I have? :)

My concern is that this is an attempt to fix something which is not broken. 
The proc cannot disappear there, which is what I believe jhb@ was pointing 
out.  What is there to fix?


More information about the cvs-src mailing list