Linus> On Sat, 10 Aug 1996, Aaron Tiensivu wrote:
>> Subject says it all.. the minor memory reorganization might have
>> been a little too overzealous.. mine locks hard after the 4k chunks
>> are displayed..
Linus> Actually, the mm fixes are correct, the bug was in just the
Linus> debugging code, so the stuff works correctly until you try to
Linus> print out the debugging info (by law of nature I found this bug
Linus> about a minute after having done the 2.0.12 release, but I
Linus> thought nobody would notice it because it doesn't show up under
Linus> any normal load, only when you press the debugging keys).
Linus> The fix is truly trivial. mm/page_alloc.c:
Linus> -#define memory_head(x) ((struct page *)x) +#define
Linus> memory_head(x) ((struct page *)(x))
Still with this fix, 2.0.12 crashes silently after doing RScroll-Lock on a
console. No reaction afterwards, no consols switching, no nothing. And no
serial console or network here at home.
But I only recompiled the kernel, I did no "make clean" after changing
mm/page_alloc.c. I don't see any reason to do so, but I'll try now. If the
error goes away that way, I'll report.
B.t.w., why do we have the need for such strange key combinations. Wouldn't
be such a report a better task for a /proc entry?
Uwe Bonnes bon@elektron.ikp.physik.th-darmstadt.de
Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------