net.c FINI & tftp stall threshold

net.c FINI & tftp stall threshold

[Seth Goldberg]

Hi,

  Is there any reason that only the following two variables’ hooks are removed at FINI time (what about the others in INIT?):

1764   grub_register_variable_hook ("net_default_server", 0, 0);
1765   grub_register_variable_hook ("pxe_default_server", 0, 0);

Also, why was the tftp stall threshold lowered to 50 (I know, it’s been that way for a while).  I ask because there is a noticeable performance impact to that smaller number for larger files (especially when running on machines with a lot of memory, that’s artificially low)?

 Thanks,
 —S

Re: net.c FINI & tftp stall threshold

[Andrey Borzenkov]

В Fri, 2 May 2014 12:00:16 -0700
Seth Goldberg <seth.goldberg@oracle.com> пишет:

> Hi,
> 
>   Is there any reason that only the following two variables’ hooks are removed at FINI time (what about the others in INIT?):
> 
> 1764   grub_register_variable_hook ("net_default_server", 0, 0);
> 1765   grub_register_variable_hook ("pxe_default_server", 0, 0);
> 

Probably none. There are quite some more places like this. I started to
collect them at some point but it turned out to be not as trivial as I
expected initially so it rather stalled.

> Also, why was the tftp stall threshold lowered to 50 (I know, it’s been that way for a while).  I ask because there is a noticeable performance impact to that smaller number for larger files (especially when running on machines with a lot of memory, that’s artificially low)?
> 
>  Thanks,
>  —S
> 
> 
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel