gmane.org.netlabs.fm2.user

Re: FM2 and traps

John Small — 2012-03-30T00:17:00

** Reply to message from "Steven Levine" <steve53-ihVZJaRskl1bRRN4PJnoQQ< at >public.gmane.org> on Sat, 11 Feb
2012 22:51:35 -0800

I finally had another couple of traps. (Details are near the end of this
message.)


After the recent traps I wrote down the specific number: 2097088K.



I never get any prompt and the system seems permanently hung.



After the first trap the dump never happened, just like before.. 

After rebooting I checked and found less than this number of bytes free on the
SADUMP partition. So I figured that the partition was too small.

I used DFSEE to expand the SADUMP drive to its maximum size (for a FAT
partition). As recommended by DFSee I then rebooted and ran "CHKDSK /f". I even
copied (and later deleted) a small file to make sure the drive was writable.

The bytes free on the drive are now 2105344000. So if the "K" in 2097088K is
1000 then I may have enough room for a dump. But if the "K" is 1024 then I do
not have enough room.

When the next trap occurred no dump was written once agai

Re: Re: FM2 and traps

2012-03-21T11:50:38

In <E1S3sFX-0000GF-00-PKVpgRpsGdXRXBi1B6Z6vp9STjOwMRefwnlmIdhGRcY< at >public.gmane.org>, on 03/03/12 
   at 08:45 AM, "Steven Levine" <steve53-ihVZJaRskl1bRRN4PJnoQQ< at >public.gmane.org> said:



OK. 
I did "prune the data" as best as I understand it. 

Let me know if...





OK. I"ll go through the mailing list when appropriate.



Done.



--
_____________________________________________________
__                                            _______
___   B e n                                     _____
____                                              ___
______                         **************      __
_____________________________________________________

Re: Re: FM2 and traps

Steven Levine — 2012-03-11T19:19:28

In <wrQPK5CrxLfq-pn2-zHNilKQrwtfd< at >dragon>, on 03/11/12
   at 03:25 PM, "Ben Dragon" <B-DragonKNOT-BZKHklv2eIvk1uMJSBkQmQ< at >public.gmane.org> said:

Hi Ben,


A bit.  Grab the current version with

  http://hobbes.nmsu.edu/h-search.php?pushbutton=Search&key=exceptq

The docs are pretty clear.  You need install two DLLs.

Steven

Re: FM2 and traps

Ben Dragon — 2012-03-11T15:25:31

On Sat, 3 Mar 2012 16:45:48 UTC, "Steven Levine" 
<steve53-ihVZJaRskl1bRRN4PJnoQQ< at >public.gmane.org> wrote:



OK. I'll do that soon.

I do not have the exceptq installed on the second machine... the one in which FM/2 crashes
more consistantly, (upon exit).

I am assuming that I need to drop the exceptq.dll in a directory that is the libpath in 
the config.sys. Is more required?

Re: Re: FM2 and traps

Steven Levine — 2012-03-03T16:45:48

In <wrQPK5CrxLfq-pn2-Ezd42LN39i7x< at >dragon>, on 03/02/12
   at 02:34 PM, "Ben Dragon" <B-DragonKNOT-BZKHklv2eIvk1uMJSBkQmQ< at >public.gmane.org> said:

Hi Ben,


Not really.  I recommend you prune the data that refers to older versions
from the logs.  It will save Gregg and I the time it takes to do this. 
When you report a defect against an older version all that will happen is
that we will you to update and verify that the defect still exists.


Gmane is a mail/news gateway.  You could subscribe directly to the fm/2
mailing list which does allow reasonable attachements.  See

 <http://svn.netlabs.org/fm2>

for instructions.  You could open a TRAC ticket and attach the relevant
files.  If you have dump files that I decide I want to see, I'll let you
know where to ftp them.

Steven

Re: FM2 and traps

Ben Dragon — 2012-03-02T14:34:24

Steven,

I collected all my FM/2 trap logs, (that I know of), including the Popup.log references.

They are attached in a ZIP file.

The "2nd-fm2_trap.log" is from a different machine and will contain the latest information
only, (I did not include that machines Popuplog).

Some of the other files contain references to older versions of FM/2 the information 
contained most likely, now being irrelevant given that updates have been made to FM/2 and 
some referenced bugs fixed. You will know better how to choose than I.

If you know or suspect, that I have other log files that I know not of, then let me know 
and I'll send those along too.


NOTE: 
GMane does not accept posts with attachments... is there an E-mail address that I can send
the zip file to?

Re: FM2 and traps

Steven Levine — 2012-02-12T06:51:35

In <201202120023.1rWrU312E3Nl36r0-yFn7Uq7GQOP+HcShpDV774oVWSWkbJJdW0UTeDyZ6EE< at >public.gmane.org>, on
02/11/12
   at 08:23 AM, "John Small" <jsmall-UxAzTYgFC9Lk1uMJSBkQmQ< at >public.gmane.org> said:

Hi John,


The exact number typically does not matter if you have 2GiB or less of
memory.  Over 2GiB is another matter.


If the dump does not complete in about a minute, it will never complete.


Normally you will get a prompt with a 30 second timeout to confirm the
dump and you can answer y or n.


That's not unexpected given what occurred.


It could be any of a number of reasons.  Keep in mind that the system is
already hosed when it is trying to set up and write the dump.


No.  Logicials are fine.


No, unless you happen to be running antique hardware that does not support
int13x.


That sounds OK.  Some 2GiB RAM setups give os2dump problems.  The dump
file includes a header and all they physical RAM.  However you really
don't have 2GiB or physical RAM.  Some of the 2GiB address space will be
used for shared vid

Re: FM2 and traps

John Small — 2012-02-11T13:23:04

** Reply to message from "John Small" <jsmall-UxAzTYgFC9Lk1uMJSBkQmQ< at >public.gmane.org> on Fri, 3 Feb 2012
06:42:24 -0500

Having set up a SADUMP partition I finally got a trap when closing FM/2.

I got a black screen telling me that a dump would be written to dive H (which
is the SADUMP partition) and that it will dump 20xxxxxx (I wish I'd written the
number down.)  bytes of memory.

I did not know what to do at that point. I waited thinking that the dump might
be in progress. But after several hours the screen never changed.

I pressed the space bar to see if it was waiting for some kind of input.
Nothing happened.

Then I remembered about CTRL+ALT+NUMLOCK+NUMLOCK and tried it. Nothing
happened. Then I tried the backspace key to "erase" the previous keystrokes
from the input buffer (if there is one at this point) followed by another
CTRL+ALT+NUMLOCK+NUMLOCK. Still nothing.

Finally I tried CTRL+ALT+DEL and the system rebooted. But there was no dump
written to H.

Can anyone tell me why no dump was written?

Re: Re: FM2 and traps

Steven Levine — 2012-02-07T20:36:21

In <wrQPK5CrxLfq-pn2-dGzKChRKxZMW< at >dragon>, on 02/07/12
   at 03:41 PM, "Ben Dragon" <B-DragonKNOT-BZKHklv2eIvk1uMJSBkQmQ< at >public.gmane.org> said:


This is a known issue.  You can ignore it.  However, we do intend to
eventually track down the reason it occurs and make it go away.

Steven

Re: FM2 and traps

Ben Dragon — 2012-02-07T15:41:50

I get this one sometimes, (but not always), when exiting FM/2, (latest version), in a 
pop-up;

"FreeCnrltem Data attempting to free 00d3b780 data twice
Module: filldir.c Line number: 1879"

I don't know if has any relationship to this issue, but maybe it will help.

Re: FM2 and traps

Steven Levine — 2012-02-07T07:40:52

In <100.981007000f8d2d4f.005-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org>, on 02/04/12
   at 12:54 PM, "Gregg Young" <ygk-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org> said:

Hi Gregg,

<isaid>
It appears you may have done an ascii upload.  The zip is corrupted and I
see some CrNl sequences in the binary.
</isaid>

The refreshed upload is clean.

I looked at pdump.302 since it appears to be the first in time.  As I
suspected, we are missing some serialization code.  The trap is at
flesh.c:213

    driveflags[*pciParent->pszFileName - 'A'] |= DRIVE_RSCANNED;

and occurs because pszFileName has been set to NULL, probably when
pciParent was released.  pszFileName was not NULL before ProcessDirectory
was called.

As we have discussed before, what needs to happen is that the shutdown
logic needs tell Flesh to quit fast and it needs to wait for the threads
to go away.  Unless we already have something equivalent, we need a
fAmShuttingDown variable.  It's probably enough for Flesh() to check
fAmShuttingDown and for the WM_C

Re: FM2 and traps

Gregg Young — 2012-02-05T04:24:43

It was done with the netdrives FTP plugin so it shouldn't have been ASCII. I just reuploaded 
it with FTPPM set to binary. Thanks

Gregg

Re: FM2 and traps

Steven Levine — 2012-02-04T22:27:35

In <100.981007000f8d2d4f.005-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org>, on 02/04/12
   at 12:54 PM, "Gregg Young" <ygk-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org> said:

Hi Gregg,


It appears you may have done an ascii upload.  The zip is corrupted and I
see some CrNl sequences in the binary.  Please check.

Thanks,

Steven

Re: FM2 and traps

Steven Levine — 2012-02-04T18:35:09

In <6B.E8.15128.D078C2F4-upiBqBNXTGrKQbo6WzeCo+cCqe+Hu3kDAL8bYrjMMd8< at >public.gmane.org>, on 02/03/12
   at 07:50 PM, msnyder1-3tLf1voIkJTQT0dZR+AlfA< at >public.gmane.org said:

Hi,


Yes.  You need

 Operating System/2 Executable File Header Utility
 Version 4.01.003 Dec 11 2003

It should be in the Toolkit shipped with recent versions of eCS.  It's
probably also available at os2site or hobbes.


Just once.  The utility modifies flags in the executables.

Steven

Re: FM2 and traps

Gregg Young — 2012-02-04T19:54:55

Steven


There are actually 3 traps before the system trap. Only one appeared in the popup log but 
based on the time stamps the other 2 occurred during the same event and that better 
matches the beep sequence that occured. I have uploaded them all. 

I think the globals exist and a semaphore locks out a lot of the UI during the scan. This 
caused the delay between my hitting close and the trap sequence. If I remember correctly 
the drive thread count dropped to 0 before the trap occurred. I am guessing I can reproduce 
this if you want me to confirm the thread count. Thanks


Gregg

Re: FM2 and traps

2012-02-04T00:50:34

On 02/03/12 at 09:25 AM, "Steven Levine" <steve53-ihVZJaRskl1bRRN4PJnoQQ< at >public.gmane.org> said:



Ahh!  My dump partition was not labled SADUMP.  I've changed it.  We'll se
whether that helps.

  exehdr /?

to get help.  The commands

This does not show a nohighmem switch.  Could it be that my version is so
old that this wasn't included?  It's dated 10-20-2000.

  exehdr /nohighmem:3 fm3.exe
  exehdr /nohighmem fm3dll.dll

Do these have to be run only once, or with each system start?

Re: FM2 and traps

2012-02-04T01:15:03

On 02/03/12 at 02:09 PM, "Gregg Young" <ygk-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org> said:



Thanks.  I installed them.  Only time will tell whether they help.  I'll
let you know.  I've opened and closed it once and it hasn't crashed, but
it doesn't always trap with the old .dll.

Re: FM2 and traps

Steven Levine — 2012-02-03T23:54:01

In <100.a8c0050018512c4f.007-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org>, on 02/03/12
   at 02:26 PM, "Gregg Young" <ygk-r9/a4iiAUwJeoWH0uzbU5w< at >public.gmane.org> said:

Hi,


I think the doscall1 trap screen is an artifact.  The cs:eip points to a
return instruction which is typical for a thread that is expected to
return from a kernel API.

The _TKDeclareInversion + 22 trap is more interesting.  I think it implies
a timing race in the kernel.  The kernel is attempting to look at a TCB
that has been erased.  The trapping instruction is

  cmp     [eax+266h], cx

and the kernel expects the eax to point to a TCB.


Please upload them.  The information is there, but it will take a bit of
poking to extract it.

FWIW, the workaround might be to delay the exit until the drive scan
thread indicates it has finished.  This will require some plumbing and a
global to tell drive scan to do a quick exit.  I don't recall how much of
this we already have in place.

Steven

Re: FM2 and traps

Gregg Young — 2012-02-03T21:26:48

Hi  Steven

I did a much better job of trapping. I hit the close button while the initial drive scan was in 
progress. All was fine until the semaphore released then FM2 trapped in DOSCALL1.DLL and 
subsequently trapped the kernel. The initial process dump isn't of much use to me at least 
as the ring 3 stack and the gate call are missing. Nothing in the kernel (actually this was a 
system trap in DOSCALL1 also) trap seem to give any indication as to what fm3dll.dll was 
doing when this all occurred. The trap screens are below. I can upload the trap dumps to 
netlabs if you would like to look at them. Thanks

Gregg

This is the initial doscall1.dll trap

Trap screen 2 found at address #70:9c84

P1=0000000b  P2=XXXXXXXX  P3=XXXXXXXX  P4=XXXXXXXX   
EAX=00000000  EBX=0000efc0  ECX=00000007  EDX=00000020 
ESI=01460a54  EDI=21920000   
DS=0053  DSACC=f0f3  DSLIM=ffffffff   
ES=0053  ESACC=f0f3  ESLIM=ffffffff   
FS=150b  FSACC=00f2  FSLIM=00000030 
GS=0000  GSACC=****  GSLIM=******** 
CS:EIP=ffef:0000e3e1  CSACC

Re: FM2 and traps

Gregg Young — 2012-02-03T21:09:27

Hi Mike

On Thu, 02 Feb 2012 10:33:21 -0500 msnyder1-3tLf1voIkJTQT0dZR+AlfA< at >public.gmane.org wrote:

I think I found the cause for this. Try the attached fm3dll.dll and let me know if it solves the 
problem. Thanks

Gregg


---------------------------------------------------------------------
To unsubscribe, e-mail: fm2-user-unsubscribe-wlNknbtBlidAfugRpC6u6w< at >public.gmane.org
For additional commands, e-mail: fm2-user-help-wlNknbtBlidAfugRpC6u6w< at >public.gmane.org

Re: FM2 and traps

Steven Levine — 2012-02-03T17:31:57

In <201202030642.1rThxc6B93Nl34l0-TXn6vOUb3+T+nIJgFVsv5bfaT9L9dGPDW08Wc4udRcs< at >public.gmane.org>, on
02/03/12
   at 06:42 AM, "John Small" <jsmall-UxAzTYgFC9Lk1uMJSBkQmQ< at >public.gmane.org> said:

Hi John,


There's no exact number for this.  The dump file consists of a header and
the physical memory image.  For a 2GiB system, the physical memory will
always less than 2GiB.  There's PCI address space and sometimes shared
video memory.


These days there's no reason not to make the dump partition 2GiB.  If you
have more than 2GiB of RAM you need to use the dumpfs version of os2dump

 <http://home.earthlink.net/~steve53/os2diags/dumpfs.txt>

Steven