Andrew Bird

Even more than that in some cases, the 'PP-DOS-GIT MFS DOSv3 share open set file attrs DOSv2' test took ~1 sec on GA (emulated) and ~12 secs on Travis(kvm)

Interestingly when I run that test here I see a kvm error, just like those I saw a while back. ~~~ ERROR: KVM_EXIT_FAIL_ENTRY: hardware_entry_failure_reason = 0x80000021 ~~~

I'll retest on devel to see if the error is specific to kvm_syn

test passes on current devel ~~~ $ test/test_dos.py PPDOSGITTestCase.test_mfs_ds3_share_open_setfattrs Test PP-DOS-GIT MFS DOSv3 share open set file attrs DOSv2 ... ok ( 3.57s) ---------------------------------------------------------------------- Ran 1 test in 3.587s OK...

> Please post output of cpuid command [cpuid.txt.gz](https://github.com/dosemu2/dosemu2/files/6337849/cpuid.txt.gz) I'll try to bisect soon.

Your latest commit still errors on my system. Bisecting now...

This 8f984bb56 seems to be the bad one for me ~~~ $ git bisect bad 8f984bb56d3ae868be9b5312205638095d8e20c5 is the first bad commit commit 8f984bb56d3ae868be9b5312205638095d8e20c5 Author: Stas Sergeev Date: Sat Feb 27...

So running with this patch on the latest kvm_syn branch ```diff diff --git a/src/base/emu-i386/kvm.c b/src/base/emu-i386/kvm.c index 1499fc25f..3c1ed7a52 100644 --- a/src/base/emu-i386/kvm.c +++ b/src/base/emu-i386/kvm.c @@ -801,12 +801,16 @@ static void kvm_regs_from_r0stack(int offs)...

Not to hijack this issue, but I wonder if these three docs might be useful/connected? [drdobbs](http://www.drdobbs.com/windows/undocumented-corner/184409317) [qlp.txt](https://www.cs.cmu.edu/~ralf/papers/qpi.txt) [gemmis.txt](https://web.archive.org/web/20170104093930/http://dgi_il.tripod.com/gemmis.txt)

And the bit about Windows taking over the EMM386 memory manager tables whilst it runs in Enhanced mode if the manager cooperates, could that be useful(not necessarily for this though)?