Back to home page

DOS ain't dead

Forum index page

Log in | Register

Back to index page
Thread view  Board view
FFK(R)

Homepage

17.12.2008, 16:25
 

HimemX v3.32 bug ? (DOSX)

I tried to use HimemX 3.32 with my 2 Go RAM laptop but mem command reports for Extended (XMS) :
Total ~ 64 MB
used ??
Free ~ 2 Go

This problem does not occur with JEMMEX 5.72 so i made a comparison between
JEMM :: _I15GetMemoryStatus and HimemX :: geti15mem and so far they seem to be equal.

Japheth(R)

Homepage

Germany (South),
17.12.2008, 16:38

@ FFK
 

HimemX v3.32 bug ?

> I tried to use HimemX 3.32 with my 2 Go RAM laptop but mem command reports
> for Extended (XMS) :
> Total ~ 64 MB
> used ??
> Free ~ 2 Go
>
> This problem does not occur with JEMMEX 5.72 so i made a comparison
> between
> JEMM :: _I15GetMemoryStatus and HimemX :: geti15mem and so far they seem
> to be equal.

what is Go? "Giga octets"? Are you French?

It's unlikely that there is a difference between HimemX and JemmEx.

All the himem's based on FreeDOS Himem have a problem with "holes" in extended memory (mostly the "os/2" hole at F00000h-FFFFFFFh). You can check with tool GETI15EX in HXRT if there is something suspicious with your memory.

---
MS-DOS forever!

FFK(R)

Homepage

17.12.2008, 19:44

@ Japheth
 

HimemX v3.32 bug ?

> what is Go? "Giga octets"? Are you French?
>
> It's unlikely that there is a difference between HimemX and JemmEx.
>
> All the himem's based on FreeDOS Himem have a problem with "holes" in
> extended memory (mostly the "os/2" hole at F00000h-FFFFFFFh). You can
> check with tool GETI15EX in HXRT if there is something suspicious with
> your memory.

Yes it's "Giga octets" which Giga bytes in english. and i'm francophone but not french.

here the result of Geti15ex.exe :
- In a Desktop PC 1 Gb RAM , HimemX 3.32 reports correctly 1 GB free and available RAM :

Int 15h, ah=88h, extended memory: 0 kB
Int 15h, ax=E801h:
ext. memory below 16 MB: 15360 (0x3C00) KB
ext. memory above 16 MB: 16116 64 KB blocks = 1007 MB [1000000-3FF3FFFF]
Int 15h, eax=E820h:
addr 000000000, size 00009FC00, type 1 (available)
addr 00009FC00, size 000000400, type 2 (reserved)
addr 0000E4000, size 00001C000, type 2 (reserved)
addr 000100000, size 03FE40000, type 1 (available)
addr 03FF40000, size 000010000, type 3 (ACPI)
addr 03FF50000, size 0000B0000, type 4 (ACPI)
addr 0FED00000, size 000000400, type 2 (reserved)
addr 0FEE00000, size 000001000, type 2 (reserved)


- In a laptop 2 Gb RAM , HimemX 3.32 reports 64 MB of free XMS and 2 GB available RAM :

Int 15h, ah=88h, extended memory: 0 kB
Int 15h, ax=E801h:
ext. memory below 16 MB: 15360 (0x3C00) KB
ext. memory above 16 MB: 30458 64 KB blocks = 1903 MB [1000000-77F9FFFF]
Int 15h, eax=E820h:
addr 000000000, size 00009FC00, type 1 (available)
addr 00009FC00, size 000000400, type 2 (reserved)
addr 0000E0000, size 000020000, type 2 (reserved)
addr 000100000, size 077EA8000, type 1 (available)
addr 077FA8000, size 000008000, type 4 (ACPI)
addr 077FB0000, size 00000E000, type 3 (ACPI)
addr 077FBE000, size 000032000, type 4 (ACPI)
addr 077FF0000, size 000010000, type 2 (reserved)
addr 0FEE00000, size 000001000, type 2 (reserved)


As you can see no memory holes on my laptop and it's exactly the same order as my desktop PC (except the additional ACPI) where HimemX works.

EDIT by rr: added {code} tag for better readability

Japheth(R)

Homepage

Germany (South),
17.12.2008, 22:22

@ FFK
 

HimemX v3.32 bug ?

> As you can see no memory holes on my laptop and it's exactly the same
> order as my desktop PC (except the additional ACPI) where HimemX works.

Yes, looks good. Now if you use the XMSSTAT tool supplied with Jemm it might give a hint what's going wrong.

---
MS-DOS forever!

FFK(R)

Homepage

18.12.2008, 01:56

@ Japheth
 

HimemX v3.32 bug ?

> Yes, looks good. Now if you use the XMSSTAT tool supplied with Jemm it
> might give a hint what's going wrong.

Rectification HimemX was not reporting correct value with mem in my desktop
with 1GB RAM and this doesn't happen with Himem.sys or XMGR.sys

Type of Memory-----Total--------Used------Free
------------------------------------------------
Conventional------655 360------84 416----570 944
Upper ------------0------------0----------0
Reserved ---------393 216------393 216 ---0
Extended (XMS)----15 728 640---?----------1 071 837 1
------------------------------------------------
Total memory------16 777 216---?----------1 072 408 1

Total under 1 MB--655 360------84 416-----570 944

Largest executable program size         570 816   (557K)
Largest free upper memory block               0     (0K)
MS-DOS is resident in the high memory area.


Here the output of XMSSTAT with HimemX on a 2GB RAM laptop:

XMS call address: 024C:061E
XMS version: 3.0
HMA handled by XMS host, HMA is allocated
largest free memory block (v2) in kB: 65535, total free: 65535
largest free memory block (v3) in kB: 1964636, total free: 1964636
XMS handle table at 024C:0030, handle cnt/size=48/10
XMS handle array at 024C:0660

 no handle region            size(kB) locks flags
--------------------------------------------------------
  1 0660   00110000-00110FFF        4     1 02 used
  2 066A   00111000-77FA7FFF  1964636     0 01 free
--------------------------------------------------------
                              1964640
free handles: 46
no UMB handler installed


and the output of XMSSTAT with Himem.sys on a 2GB RAM laptop :

XMS call address: 024C:00CF
XMS version: 3.0
HMA handled by XMS host, HMA is allocated
largest free memory block (v2) in kB: 65535, total free: 65535
largest free memory block (v3) in kB: 1964636, total free: 1964636
XMS handle table at FFF8:CC90, handle cnt/size=48/10
XMS handle array at FFF8:D58E

 no handle region            size(kB) locks flags
--------------------------------------------------------
  1 D58E   00110000-00110FFF        4     1 02 used
  2 D598   00111000-77FA7FFF  1964636     0 01 free
--------------------------------------------------------
                              1964640
free handles: 46
no UMB handler installed


EDIT by rr: added {code} tag for better readability

Japheth(R)

Homepage

Germany (South),
18.12.2008, 21:16

@ FFK
 

HimemX v3.32 bug ?

Ok, but XMSSTAT reveals that there is no problem with HimemX. It's a MEM problem.

---
MS-DOS forever!

FFK(R)

Homepage

19.12.2008, 10:44

@ Japheth
 

HimemX v3.32 bug ?

> Ok, but XMSSTAT reveals that there is no problem with HimemX. It's a MEM
> problem.

I agree that it's not a critical bug. But on all JEMMEX, Ms HIMEM.SYS and XMGR : MEM reports correct values. So it's a HimemX minor bug i guess.

Japheth(R)

Homepage

Germany (South),
19.12.2008, 13:16

@ FFK
 

HimemX v3.32 bug ?

> I agree that it's not a critical bug. But on all JEMMEX, Ms HIMEM.SYS and
> XMGR : MEM reports correct values. So it's a HimemX minor bug i guess.

I just noticed that in the tests with XMSSTAT you tested the 2GB machine only but not the 1GB desktop machine which is the interesting one. Why?

---
MS-DOS forever!

FFK(R)

Homepage

23.12.2008, 21:23

@ Japheth
 

HimemX v3.32 bug ?

> > I agree that it's not a critical bug. But on all JEMMEX, Ms HIMEM.SYS
> and
> > XMGR : MEM reports correct values. So it's a HimemX minor bug i guess.
>
> I just noticed that in the tests with XMSSTAT you tested the 2GB machine
> only but not the 1GB desktop machine which is the interesting one. Why?

but in both 1Gb and 2Gb machines "MEM" report bad amount of memory

here the XMSSTAT of the 1gb machine :

XMS call address: 025C:061E
XMS version: 3.0
HMA handled by XMS host, HMA is allocated
largest free memory block (v2) in kB: 65535, total free: 65535
largest free memory block (v3) in kB: 1046716, total free: 1046716
XMS handle table at 025C:0030, handle cnt/size=48/10
XMS handle array at 025C:0660

 no handle region            size(kB) locks flags
--------------------------------------------------------
  1 0660   00110000-00110FFF        4     1 02 used
  2 066A   00111000-3FF3FFFF  1046716     0 01 free
--------------------------------------------------------
                              1046720
free handles: 46
no UMB handler installed


with mem i get :
Total memory:   Total=16 384K        Used=?   Free=943 422K

EDIT by rr: added {code} tag for better readability

DOS386(R)

29.12.2008, 01:59

@ FFK
 

HimemX v3.32 bug ? DUGL

FFK wrote:

> use HimemX 3.32 with my 2 Go RAM laptop but mem command reports

Missed something ? http://dugl.50webs.com/ :-)

---
This is a LOGITECH mouse driver, but some software expect here
the following string:*** This is Copyright 1983 Microsoft ***

Rugxulo(R)

Homepage

Usono,
29.12.2008, 23:01

@ DOS386
 

HimemX v3.32 bug ? DUGL

> FFK wrote:
>
> > use HimemX 3.32 with my 2 Go RAM laptop but mem command reports
>
> Missed something ? http://dugl.50webs.com/ :-)

You mean the new DUGL 1.13? (No mention of new DuglPlus, though.)

P.S. DuglPlayer really should be recompiled with latest libMPEG2 since the latter now has optional SSE2 support, which latest CWSDPMI allows (unless it's compile-time only, ugh).

DOS386(R)

30.12.2008, 22:53

@ Rugxulo
 

HimemX v3.32 bug ? DUGL

> You mean the new DUGL 1.13?

YES.

> DuglPlayer really should be recompiled with latest libMPEG2 since the
> latter now has optional SSE2 support, which latest CWSDPMI allows

HDPMI32 used to for years ... and there is the great D3X Ring0 extender :-)

---
This is a LOGITECH mouse driver, but some software expect here
the following string:*** This is Copyright 1983 Microsoft ***

FFK(R)

Homepage

08.01.2009, 15:50

@ DOS386
 

HimemX v3.32 bug ? DUGL

> > DuglPlayer really should be recompiled with latest libMPEG2 since the
> > latter now has optional SSE2 support, which latest CWSDPMI allows
>
> HDPMI32 used to for years ... and there is the great D3X Ring0 extender
> :-)

Does D3X Ring0 extender enable SSE ?

Rugxulo(R)

Homepage

Usono,
09.01.2009, 00:48

@ FFK
 

D3X extender + SSE

> Does D3X Ring0 extender enable SSE ?

No, but at ring 0 you can do it yourself. (Or run some other app first, it only needs to be done once each session.) I meant to eventually write a patch for D3X but haven't done it yet.

---
Know your limits.h

FFK(R)

Homepage

31.12.2008, 09:34

@ Rugxulo
 

HimemX v3.32 bug ? DUGL

> You mean the new DUGL 1.13? (No mention of new DuglPlus, though.)

Yes DUGL 1.13 is out with a little introduced bug an 8bpp Poly on the double sided flag, but who care ?
for DuglPlus except a new Sbpro driver and 16bpp port of the GUI there will be no additional functionnality

> P.S. DuglPlayer really should be recompiled with latest libMPEG2 since the
> latter now has optional SSE2 support, which latest CWSDPMI allows (unless
> it's compile-time only, ugh).

For DuglPlayer i do wished to get rid of the MpgSplit step before playing any mpeg but I don't have enough knowledge and time to learn how to demux and no one want to contribute :(
Any way if you have the djgpp port of the new libMPEG2 i will recompile DuglPlayer and provide a newer version.

Rugxulo(R)

Homepage

Usono,
02.01.2009, 05:01

@ FFK
 

HimemX v3.32 bug ? DUGL

> For DuglPlayer i do wished to get rid of the MpgSplit step before playing
> any mpeg but I don't have enough knowledge and time to learn how to demux
> and no one want to contribute :(

No idea, sorry.

> Any way if you have the djgpp port of the new libMPEG2 i will recompile
> DuglPlayer and provide a newer version.

I haven't looked at it yet, was just thinking outloud. :-P

FFK(R)

Homepage

08.01.2009, 15:49

@ Rugxulo
 

HimemX v3.32 bug ? DUGL

> > Any way if you have the djgpp port of the new libMPEG2 i will recompile
> > DuglPlayer and provide a newer version.
>
> I haven't looked at it yet, was just thinking outloud. :-P

It seems that libmpeg2 isn't hard to compile for DJGPP but i do prefer waiting that some one compile it for us :-) Until i finish porting the DUGL-Plus GUI to 16bpp so then DUGL Player will be faster and source code easier to understand.

Back to index page
Thread view  Board view
15112 Postings in 1359 Threads, 247 registered users, 11 users online (0 registered, 11 guests)
DOS ain't dead | Admin contact
RSS Feed
powered by my little forum