DOS ain't dead

The QEMU team has released version 0.10.4 on 12 May 2009.

Home page: http://www.qemu.org/
Download (sources): qemu-0.10.4.tar.gz
Announcement: ANNOUNCE: Release 0.10.4 of QEMU

Changes for version 0.10.4:
- Improve block range checks to remove integer overflow (Kevin Wolf)
- e1000: do not re-init PCI config space 0 (Amit Shah)
- fix AIO deletion race (Alex Graf)
- reset option roms on reboot (Glauber Costa)
- fix qcow2 corruption in cluster freeing (Gleb Natapov)
- Enable power button event generation (Gleb Natapov)

Changes for version 0.10.3:
- fix AIO cancellations (Avi Kivity)
- fix live migration error path on incoming
- avoid SEGV on pci hotplug failure (Chris Wright)
- fix serial option in -drive
- support DDIM for option roms (Glauber Costa)
- avoid fork/exec on pre-2.6.27 kernels with KVM (Jan Kiszka)
- block-vpc: don't silently create smaller images than requested (Kevin Wolf)
- Fix non-ACPI timer interrupt routing (Beth Kon)
- hpet: fix emulation of HPET_TN_SETVAL (Jan Kiszka)
- kvm: fix cpuid initialization (Jan Kiszka)
- qcow2: fix corruption on little endian hosts (Kevin Wolf)
- avoid leaing memory on hot unplug (Mark McLoughlin)
- fix savevm/migration after hot unplug (Mark McLoughlin)
- Fix keyboard mapping on newer Xords with non-default keymaps (balrog)
- Make PCI config status register read-only (Anthony Liguori)
- Fix crash on resolution change -> screen dump -> vga redraw (Avi Kivity)

Because I'm still busy with moving to new rooms, I wish to point you to some other web sites for recent Win32 binaries.

1) http://qemu-forum.ipi.fi/viewtopic.php?f=5&t=5028 (postings by user 'zfeet')
2) http://sourceforge.net/projects/winqemu/ (MS VC++ 2008)
3) http://lassauge.free.fr/qemu/

---
Forum admin

> The QEMU team has released version 0.10.4 on 12 May 2009.

You mentioned before that you also used GCC 4.3.0, but I've since found an unofficial 4.4.0 for MinGW that seems to work pretty well:

http://www.tdragon.net/recentgcc/

> You mentioned before that you also used GCC 4.3.0, but I've since found an
> unofficial 4.4.0 for MinGW that seems to work pretty well:
>
> http://www.tdragon.net/recentgcc/

I'm aware of this build, but as it's unsupported by the MinGW guys, I didn't want to use it.

---
Forum admin

> You mentioned before that you also used GCC 4.3.0, but I've since found
> an unofficial 4.4.0 for MinGW that seems to work pretty well:

Sherpya has GCC 4.5.0 , but don't ask me where he got it ... apparently NOT from MinGW page :

cd /d c:\MinGW
gzip -df binutils-2.19.1-mingw32-bin.tar.gz
tar -xf binutils-2.19.1-mingw32-bin.tar
cd info
ren dir dir.binutils
cd ..
gzip -df gcc-core-3.4.5-20060117-3.tar.gz
tar -xf gcc-core-3.4.5-20060117-3.tar
cd info
copy dir dir.gcc
copy dir+dir.binutils
cd ..
gzip -df gcc-g++-3.4.5-20060117-3.tar.gz
tar -xf gcc-g++-3.4.5-20060117-3.tar
gzip -df mingwrt-3.15.2-mingw32-dev.tar.gz
tar -xf mingwrt-3.15.2-mingw32-dev.tar
gzip -df mingwrt-3.15.2-mingw32-dll.tar.gz
tar -xf mingwrt-3.15.2-mingw32-dll.tar
gzip -df w32api-3.13-mingw32-dev.tar.gz
tar -xf w32api-3.3-mingw32-dev.tar

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

> > You mentioned before that you also used GCC 4.3.0, but I've since found
> an
> > unofficial 4.4.0 for MinGW that seems to work pretty well:
> >
> > http://www.tdragon.net/recentgcc/
>
> I'm aware of this build, but as it's unsupported by the MinGW guys, I
> didn't want to use it.

Well, they don't support anything besides 3.4.5, right? So even the 4.3.0 is risking it. Anyways, I'd assume 4.4.0 is more stable than 4.3.0 (esp. since 4.3.3 exists).

BTW, the two advantages to "just testing" with 4.4.0 is: a). you can see if it improves performance, and b). you can see if anything breaks or acts weird with 4.4.0 (which DJGPP still lacks). For instance, I tested 4.3.0 before it was official, and discovered their silly "-Wparentheses now enabled in -Wall" decision. Sometimes it's good to know these things as soon as possible!!