home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!beartrk!ceilidh!hijo-2!dnichols
- From: dnichols@ceilidh.beartrack.com (Don Nichols (DoN.))
- Newsgroups: comp.sys.3b1
- Subject: Re: Ping for TCP/IP
- Keywords: raw ethernet, ping, read, write
- Message-ID: <1992Aug29.003521.16561@ceilidh.beartrack.com>
- Date: 29 Aug 92 00:35:21 GMT
- References: <714157972@macbeth.cs.duke.edu> <1992Aug28.143120.25887@jhunix.hcf.jhu.edu> <715021719@macbeth.cs.duke.edu>
- Organization: D and D Data, Vienna Virginia
- Lines: 45
-
- In article <715021719@macbeth.cs.duke.edu> pusateri@duke.cs.duke.edu (Thomas Pusateri) writes:
- >In article <1992Aug28.143120.25887@jhunix.hcf.jhu.edu>
- >andy@jhunix.hcf.jhu.edu (Andy Poling) writes:
- >>
- >>I too spent some time on ping (a long time ago). My recollection is the
-
- As have I. Is there anyone left who hasn't? :-)
-
- >>same - I could never successfully read raw packets. I don't think the
- >>SOCK_RAW socket support is complete...
- >
- >If we are ever going to solve problems like this, we are going to need either
- >a way to recompile the kernel source and put printf statements in it
- >or some type of kernel debugger to set break points and print registers
- >and assembly code.
-
- There is a kernel debugger, I think still in the archives in
- OSU-CIS, and there are hooks to it in the 3.51m kernel, if I remember
- correctly. I think that is was simply called kdb, which was also suggested
- as meaning "kernel de buggerer" in the wrong hands. :-)
-
- Another consideration is that access to the kernel sources, when and
- if it occurs, may not solve our problems with the ethernet, since the
- ethernet is using a loadable device driver from a third party, and *that*
- source may not be available to us.
-
- >Even if we can get source some day, it would still be nice to have a kernel
- >debugger or monitor. Does anyone at AT&T or convergent know if there are
- >any development tools that could be released as well under the possible
- >upcoming source agreement?
-
- Yes, anything else would be a welcome addition. I'm not sure how
- much use kdb is without the source access, so we know what we're looking at,
- instead of guessing, based on the assumption that meaningful names were
- squeezed into 8-character variable names. (At least the earlier versions of
- the compiler only kept a limited subset of the variable names as
- significant, so that probably is reflected in the kernel code, even if the
- most recent parts were compiled with the newer compiler.) After all, who
- goes back and expands variable names in working code? :-)
-
- --
- Donald Nichols (DoN.) | Voice (Days): (703) 704-2280 (Eves): (703) 938-4564
- D&D Data | Email: <dnichols@ceilidh.beartrack.com>
- I said it - no one else | <nichols@nvl.army.mil>
- --- Black Holes are where God is dividing by zero ---
-