Win7 where is the PSW (CS:IP) of a thread on the ready Q in a BSOD?

[MtheK]

  I found 'INVOKE FlushFileBuffers' which I now do B 4 the 'INVOKE
CloseHandle' (files only), which seems to not only have stopped the
null-corruption of my non-FILE_FLAG_NO_BUFFERING files, but, now,
even for the .BAT ECHOs? Where I was able to re-create the corruption
at will by waiting 15 seconds w/the simple .BAT ECHO, now, after my
program changes, I can't. Strange. I was hoping to find a "commit"
command I could use in the .BAT btx the ECHO and the GOTO loop, but,
maybe, I now won't have to?

  For my normal file, I now do it after writing every 100 records,
and it seems to have eliminated any corruption at all (possibly, now,
just missing data (in my case, only up to apx 5K)).  My respect for
this OS has gone up a notch...

[MtheK]

  Nice to end the week on a positive note B 4 the big racing marathon:
F1, INDY, NASCAR, NHRA.

  On my 4th BSOD since, and a few after (14 since I started this),
the simple .BAT O/P again got null-corrupted. The DSNTODAY *AB .BAT
ECHOs also may survive intact, or may not. I changed my TOUCH program
to flush the buffers B 4 the close, and will insert it btx the ECHO
and the start of the GOTO loop, as a "commit", and re-try a few more
times, to see if it ever gets corrupted again...
 
  Meanwhile, all my files in my programs that use either
FILE_FLAG_NO_BUFFERING or 'INVOKE FlushFileBuffers' now survive intact,
with no null-corruption. My normal file lost only 47 records compared
to memory, within the 100 limit I set for it. Vastly superior to a
half-M file of all nulls.
 
  Perhaps 'CacheFileTimeout' in the registry? Not sure if this is
network-only...

[MtheK]

  Split off to here:

http://masm32.com/board/index.php?topic=4274.0