scruty - 2010-10-26

Hello,
The CVS build is running much better than latest snapshot (20080503_1030).

I'm using OpenVMS 8.3 latest cvs build on Ubuntu 10.04, using putty (telnet), logged-in on two consoles. (the network (TCP/IP) is up on the es40 side. but I don't yet know what must be done network-wise on the host side. anyway, I don't need the network currently.)

Except for some (mostly S-)floating point issues, all of my many own test programs I tried so far now work. I also heavily used the linker without problems. The analyze/disk/repair now no longer reports strange errors and was even able to fix the few which were correctly reported. Only once in the middle of a long session I got an access violation (see below), but when I retried, it worked.

Only once during a long test session I got an access violation*, but when I retried immediately, it worked.

I strongly believe the fix reported in http://www.es40.org/Project_Status
"16-MAR-2009: David Hittner has fixed a problem in the CPU's handling of super-page bits."
helped a lot to get rid of these access violations**. Thank you! Maybe there's another 'malvolent' bit hiding somewhere.
So please David, if you have an idea…

I'll try to fix the floating the few remaining point errors and then run more tests programs.

Best regards,
scruty

*) 

$ dir *.com.
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=FFFFFFFFF
FFFFFB8, PC=000000007AFA1E00, PS=0000001B
  Improperly handled condition, image exit forced.
    Signal arguments:   Number = 0000000000000005
                        Name   = 000000000000000C
                                 0000000000000000
                                 FFFFFFFFFFFFFFB8
                                 000000007AFA1E00
                                 000000000000001B
    Register dump:
    R0  = 000000007AF08548  R1  = 000000007ADF3648  R2  = 000000007AF08618
    R3  = 000000007ADF36C0  R4  = 00000000000310FC  R5  = 000000007FFEFD00
    R6  = 000000007ADF3668  R7  = 000000007AE0AE9C  R8  = 000000007FF9CDE8
    R9  = 000000007ADF38B0  R10 = 000000007ADF36C0  R11 = 0000000000000001
    R12 = 000000007ADF3A98  R13 = 000000007AF0FFD8  R14 = 0000000000020000
    R15 = 000000007AF0C660  R16 = 000000007ADF3668  R17 = 0000000000000000
    R18 = 000000007ADF3610  R19 = 0000000000000000  R20 = 0000000000000002
    R21 = 000000007ADF364E  R22 = 0000000000000006  R23 = 000000007ADF361A
    R24 = 0000000000000000  R25 = 0000000000011228  R26 = 000000007AFA1DF4
    R27 = 000000007AF08520  R28 = 0000000000000001  R29 = 000000007ADF3640
    SP  = 000000007ADF3640  PC  = 000000007AFA1E00  PS  = 000000000000001B
$

**) it's one of those access violation I talked about in my project review here, 2 weeks ago:
http://sourceforge.net/projects/es40/reviews/