[9front] [PATCH] remove misleading BUGS from auth(8)

[9front] [PATCH] remove misleading BUGS from auth(8)

[Jacob Moody]

I believe this "bug" is no longer accurate, seems like you
can change userid to none even on a terminal. I also would
expect a terminal to permit changing to any userid that is
present in the hostowner factotum, but wanted to double
check in case I am missing something.


Thanks,
moody

---
--- a//sys/man/8/auth
+++ b//sys/man/8/auth
@@ -299,5 +299,3 @@
 .IR authsrv (2),
 .IR keyfs (4),
 .IR securenet (8)
-.SH BUGS
-Only CPU kernels permit changing userid.

Re: [9front] [PATCH] remove misleading BUGS from auth(8)

[mkf9]

Jacob Moody wrote:
> I believe this "bug" is no longer accurate, seems like you
> can change userid to none even on a terminal. I also would
> expect a terminal to permit changing to any userid that is
> present in the hostowner factotum, but wanted to double
> check in case I am missing something.

do different $services use different kernels in 9front
or it's a legacy of bell labs?

Re: [9front] [PATCH] remove misleading BUGS from auth(8)

[Jacob Moody]

On 6/2/22 10:28, mkf9 wrote:
> Jacob Moody wrote:
>> I believe this "bug" is no longer accurate, seems like you
>> can change userid to none even on a terminal. I also would
>> expect a terminal to permit changing to any userid that is
>> present in the hostowner factotum, but wanted to double
>> check in case I am missing something.
> 
> do different $services use different kernels in 9front
> or it's a legacy of bell labs?
> 

In 9front its all one kernel with different code paths depending on $service.
 From my understanding of the code this is mostly to set resource defaults,
and different init scripts (/bin/termrc vs /bin/cpurc).
If my history serves me correct they did used to actually be different
builds. The wording here does give away how old this "bug" is. In any
case, this section of the man page only serves to further confuse things.
I've committed the change.


Thanks,
moody

Re: [9front] [PATCH] remove misleading BUGS from auth(8)

[Stanley Lieber]

9front just has one kernel for everything.

sl

> On Jun 2, 2022, at 12:29 PM, mkf9 <mkf9@riseup.net> wrote:
> 
> Jacob Moody wrote:
>> I believe this "bug" is no longer accurate, seems like you
>> can change userid to none even on a terminal. I also would
>> expect a terminal to permit changing to any userid that is
>> present in the hostowner factotum, but wanted to double
>> check in case I am missing something.
> 
> do different $services use different kernels in 9front
> or it's a legacy of bell labs?
> 
> 

Re: [9front] [PATCH] remove misleading BUGS from auth(8)

[mkf9]

Jacob Moody wrote:
>> do different $services use different kernels in 9front
>> or it's a legacy of bell labs?
>>
> 
> In 9front its all one kernel with different code paths depending on $service.
>   From my understanding of the code this is mostly to set resource defaults,
> and different init scripts (/bin/termrc vs /bin/cpurc).
> If my history serves me correct they did used to actually be different
> builds. The wording here does give away how old this "bug" is. In any
> case, this section of the man page only serves to further confuse things.
> I've committed the change.
> 
> 
> Thanks,
> moody
> 

I see, thank you.