Re: Re: /proc file system, seems to -not- have standardisation ?

Mr. James W. Laferriere Network Engineer (babydr@nwrain.net)
Sun, 6 Apr 1997 14:08:07 -0700 (PDT)


OK, Now after being -extremely- sick, then my OS HD going toast.

Mr. "Theodore Y. Ts'o" <tytso@MIT.EDU> , Suggested that defining
the Trouble being seen is a better approach. OK. ;-)
(Ted, I am somewhat freely interpreting your previous message)

------------------------------------------------------------------------

1) When I look into /proc/scsi I see the name of the driver(s) in
use. & a file that has info about scsi devices. (Not that I am
condoning / disparaging any of the below methods of presenting
information.) IE:

a) /proc/scsi/ncr53c8xx < dir to info about the driver.
/proc/scsi/scsi < info about the devices found.
/proc/scsi/ncr53c8xx/0 < info about ctrl 0 driver.
/proc/scsi/ncr53c8xx/1 < info about ctrl 1 driver.

When I look into /proc/net I see a whole bunch of files
which have a lot of info about the network , but nothing
concerning the (device) driver. IE:

b) /proc/net/aliases < NET device aliasing
/proc/net/alias_types < Types of aliasing availale
/proc/net/dev < various info per dev & type
/proc/net/rt_cache < Routing cache
/proc/net/route < present routing table
/proc/net/udp < Processes/? using UDP
/proc/net/tcp < Processes/? using TCP
/proc/net/sockstat < Socket use statistics
/proc/net/snmp < snmp MIB's in use & stats
/proc/net/raw < Processes/? using RAW Socket
/proc/net/rarp < Reverse ARP table
/proc/net/ip_mr_cache < Multicast Routing cache ?
/proc/net/ip_mr_vif < Multicast Routing Virtual inferface ?
/proc/net/igmp < Inter Gateway Meta(?) protocol
/proc/net/arp < The ARP table.
/proc/net/unix < UNIX Sockets open.

And the next directory in proc I'd have gone to concerning
networks is /proc/sys/net , But it seems to be mostly empty.

bb) /proc/sys/net/802/ < MT Dir.
/proc/sys/net/core/ < MT Dir.
/proc/sys/net/ethernet/ < MT Dir.
/proc/sys/net/ipv4/ < The only one with info.
/proc/sys/net/ipv4/arp_check_interval
/proc/sys/net/ipv4/arp_confirm_interval
/proc/sys/net/ipv4/arp_confirm_timeout
/proc/sys/net/ipv4/arp_dead_res_time
/proc/sys/net/ipv4/arp_max_tries
/proc/sys/net/ipv4/arp_res_time
/proc/sys/net/ipv4/arp_timeout
/proc/sys/net/ipv4/ip_forward
/proc/sys/net/unix/ < MT Dir.

2) 1a-1bb) Summation of Concerns.

a) In these directories there should be information
concerning the device driver from which it is gleaning
the other info from. IE: Could be just a pointer.

b) What I hope could be used.

for:
1a) /proc/scsi/ncr53c8xx-0 < dir to info about ctrl 0 driver.
/proc/scsi/ncr53c8xx-0/ctrl
< info about ctrl 0 driver.
/proc/scsi/ncr53c8xx-0/devices
< info about devices on ctrl 0 .
/proc/scsi/ncr53c8xx-0/?????
< further info about devices &/or ctrl 0 .

(If scsi0 (from dmesg) is ncr53c8xx-0 then a link or pointer back
to scsi0 may be in order.)

1b-bb) /proc/net/eepro100-0 < dir to info about ctrl 0 driver.
/proc/net/eepro100-0/ctrl
< info about ctrl 0 driver.
/proc/net/eepro100-0/????
< info about ???? (see 1b & 1bb above) .

(If eth0 (from dmesg) is eepro100-0 then a link or pointer back
to eth0 may be in order.)

------------------------------------------------------------------------

These concerns are what prompted the previous flurry of messages
concerning this topic. Please RSVP. the suggestion in 2b) has
had a great deal of thought put into it & is based on several
converstions both on & off the list.

Tia, JimL
+-----------------------------------------------------------------------+
| James W. Laferriere - Network Engineer - babydr@nwrain.net |
| System Techniques - 25416 - 22nd S. - Kent, WA 98032 |
| Give me VMS -or- Give me Linux -but- only on AXP |
+-----------------------------------------------------------------------+
|-> Linux-Vax Port, Still in Progress . IE: No Progress To Report ;-) <-|
+-----------------------------------------------------------------------+