[Date Prev][Date Next][Subject Prev][Subject Next][
Date Index][
Subject Index]
Re Keystrokes: Possible BIOS compatibility
- Subject: Re Keystrokes: Possible BIOS compatibility
- From: "Patricia M. Godfrey" PriscaMG@xxxxxxxx
- Date: Thu, 28 Jul 2005 13:28:43 -0400
Robert suggested testing again on the fictitious functions RQ and UV.
Now things get really weird. I still had a sneaking suspicion that OOM
was part of the problem. So I modified Robert's debugging version to
test only functions beginning with M, N, R, and U (see below). On the
first go-round, it crashed on NL with the familiar message: Illegal
operation: invalid instruction, fault location 0000:0017. So I included
NL in the omitted items, and tried again. This time it did not crash at
RQ or UV nor had it crashed, or did it crash, at NI, which was the
original precipitating call.
So I tried again, with the whole megillah and only NL excluded. Ran OK.
One oddity: my startup.int includes AutoSave settings: default aot=1,5.
During the first few minutes, I would see the "autosaving to tmp file"
message flash below the command line. Later on I didn't, but it
apparently was saving, since AUTOSAV2.TMP contains everything through YD.
I could NOT have mistaken a spontaneous reboot for anything else. It
happened. Twice. Why is the question. And why it didn't this time. It
occurs to me that some misbehaving Windows app may have glommed onto
some RAM the other day and not let go. I hadn't run anything else today,
except Thunderbird and Juno. On the other hand, surely a spontaneous
reboot would have cleared RAM?
In any event, NL is evidently a trapping call on this machine. It might
be helpful if we could see which BIOSes trap which calls. This is, as I
may have mentioned, an AMI BIOS. Except for the 2 Compaqs and the IBM
Thinkpad (will test the Compaq laptop next, but that is a pathetically
underpowered machine: 24 MB of RAM), all of my machines are. Would
anyone with an Award or Phoenix BIOS like to try? (We do all know what
BIOS is on our machines, no? Or how to find out? I can post
instructions, if the Wise Persons think it might be helpful.)
Patricia M. Godfrey
XPLeNCODE v2.0
b-gin [UNTITLED]
{<}GLC{>}{{;5keystrokes}} List current .KBD assignments for f
unctions RJH LastRev.7/26/05 -- *Debugging Version* -- RUN a
s freestanding PM![cr|lf]{002}{<}LBC{>}{<}SX01,{<}VA$WA{>}{>}
{<}IF{<}PV01{>}<1{>}{<}PRNo window available{>}{<}EX{>}{<}EI{
>}[BX_]func #{<}PV01{>}[Q2_][BX_]ne[Q2_][BX_]wait[Q2_][XP_]{<
}SV01,MNRU{>}{<}SV02,0123456789@ABCDEFGHIJKLMNOPQRSTUVWXYZ{>}
{<}SV03,{027}X{>}{<}SU04,{<}GT05{>}{<}GT06{>}: {<}GT07{>}[255
+048+068]{>}{<}LBA{>}{<}SX08,{<}IS01{>}{>}{<}SX09,{<}IS02{>}{
>}{<}XS08,03,,05,01{>}{<}LBB{>}{<}XS09,03,,06,10{>};*;[cr|lf]
;*;[cr|lf]{<}IF("|"+{<}IS05{>}+{<}IS06{>}+"|"){238}"|NL|"<0{>
};*; <==Add 2-char funcs to EXCLUDE[cr|lf];*; here, in format
"|MS|AA|BB|CC|"etc.[cr|lf];*;[cr|lf];*; Starting with "N{sec
ond_character}", display 2 chars on the CMline,[cr|lf];*; the
n wait one second before testing for validity of the two[cr|l
f];*; characters with VA_ :[cr|lf]{<}IF{<}IS05{>}>"L"{>}[BC_]
{<}PV05{>}{<}PV06{>}[GT_][BX_]p 1[Q2_]{<}EI{>};*;[cr|lf];*; T
o revert to FAST processing, comment out (with leading ";*;")
[cr|lf];*; the line above[cr|lf];*;[cr|lf]{<}SX11,"{<}SX07,{<
}VA_"+{<}IS05{>}+{<}IS06{>}+"{>}{>}"{>}{<}PV11{>}{<}IF{<}VA|0
7{>}>0{>}{<}GT04{>}{<}EI{>}{<}EI{>}{<}IF{<}VA|10{>}>0{>}{<}SX
09,{<}IS10{>}{>}{<}GLB{>}{<}EI{>}{<}IF{<}VA|01{>}>0{>}{<}GLA{
>}{<}EI{>}[TF_]{<}PRDone{>}{<}EX{>}{002}[cr|lf][cr|lf]
-nd
XPLeNCODE