Heise reports from SCO Forum
Posted Aug 20, 2003 8:52 UTC (Wed) by
keithl (guest, #14228)
Parent article:
Heise reports from SCO Forum
Non programmer here. I looked at ftp.kernel.org for the ate_utils.c
file. I note that ate_utils.o appears in .../arch/ia64/sn/io/Makefile
between 2.4.3 (March 30, 2001) and 2.4.4 (April 28, 2001). The
ate_utils.c file itself shows up in that directory between 2.4.18
(Feb 25, 2002) and 2.4.19 (Aug 2, 2002), complete. The routine
shown in the article, atealloc(), shows up in 2.4.4 as a single
define in .../arch/ia64/sn/io/pcibr.c:
#ifdef __ia64
#define rmallocmap atemapalloc
#define rmfreemap atemapfree
#define rmfree atefree
#define rmalloc atealloc
#endif
While in 2.4.19 it is in a bunch of places, still associated with
variations of pcibr.c. Perhaps someone with far more code
understanding than I can explain what all this code is for, and
how the code got in there, who wrote it, and why the .o is
mentioned in the 2.4.4 makefile over a year before the actual
code arrives, and maybe even how the rmalloc [that is, atealloc]
works when atealloc isn't there yet?
No agenda here, I'm just an egg wondering how all this works, so I
can pontificate at length to the two or three folks out there who
know even less than I do. Perhaps some kernel savvy individual
might inform us what this all means. The cool thing about Linux
is that almost all of this is documented in public somewhere, if
you know where to look, and some individual is always ready with
a clue-by-four to enlighten the rest of us.
Keith
(
Log in to post comments)
