[Toybox] [TOY] lspci

[Rob Landley]

On 07/27/2013 02:56:53 AM, Felix Janda wrote:
> Rob Landley wrote:
> > Did I miss an intermediate patch? Could you send me a fresh one that
> > applies to my current tree?
> 
> In the current tree I see that you have applied a newer version  
> anyway.
> (with some minor cleanup of yours.)

I've been slowly catching up on email. Currently only about 2000  
messages behind. (Mostly linux-kernel and qemu-devel stuff.)

> Unfortunately that version has "lspci -e" broken as observed by Isaac
> somewhere in this thread. Below is a fix. (One could also replace the
> "!=" by "==".)

Applied.

> Anyway Isaac is doing (or planning to do) a major rewrite enabling  
> lspci
> to read the pciid database for human readable names of vendors and  
> devices.

I look forward to it.

Ashwini Sharma sent a "ping" implementation from his guys, but between  
my discomfort at applying 3-clause BSD code (it's not the same license)  
and the fact it shelled out to "ping6" to do ipv6 support instead of  
having it unified, I decided to just do my own.

It's a learning experience. I'm reading the code he sent and the  
busybox implementation and running the host version under strace to try  
to figure out what the heck ping _does_. Plus reading various  
structures and going "the ipv6 version and ipv4 version are the exact  
same thing followed by a union and the only thing that differs is the  
fields in the union, so why the heck did they fork this mess instead of  
having _one_?" Plus it's not an error to open an AF_INET6 socket with  
IPPROTO_ICMP instead of IPPROTO_ICMPV6 and I'm looking in the kernel to  
see if that's just a coincidental collision of constants or what... I'd  
like to reduce gratuitous duplication. Possibly I can put all these  
constants in a table and select v6 vs v4 out of the table...

Anyway, learning curve. Eating more time than I expected.

Thanks,

Rob