1 files changed, 54 insertions, 0 deletions
diff --git a/results/classifier/zero-shot/108/other/616769 b/results/classifier/zero-shot/108/other/616769
new file mode 100644
index 00000000..c5a47f43
--- /dev/null
+++ b/results/classifier/zero-shot/108/other/616769
@@ -0,0 +1,54 @@
+graphic: 0.711
+performance: 0.691
+semantic: 0.660
+device: 0.634
+debug: 0.567
+PID: 0.533
+permissions: 0.524
+network: 0.524
+other: 0.502
+boot: 0.456
+socket: 0.453
+vnc: 0.424
+KVM: 0.299
+files: 0.287
+
+interrupt problem x86_64
+
+Hello.
+I'm studing the x86_64 arch to port colinux to this new architecture.
+
+For who does not know, colinux is a port of linux that runs on windows NATIVELY. Colinux is
+1) a small windows driver
+2) a kernel patched
+3) some windows user-space application that talk with linux kernel (network, console, ...)
+
+I have written the more internal part of colinux, that is the code that switch between windows and linux.
+This is done saving and restore the machine state :
+1) CR3
+2) IDT
+3) registers
+
+The problem I see is that after the switch I see the reboot of my virtual machine.
+I would say that the new IDT and/or CR3 is not flushed.
+My code is written in asm/C so I can follow the code step by step.
+I can say exactly the instruction that is broken.
+
+All my code runs nicely on bochs.
+I don't have an x86_64 real PC.
+
+If someone wants to repair this bug .... I'm here.
+
+Paolo Minazzi
+
+Which version of QEMU did you use when you encountered this problem? Can you still reproduce this issue with the latest version of QEMU?
+
+Hello Thomas,
+it happens some years ago and now I'm not able to reproduce it.
+I'm sorry,
+thanks again,
+Paolo
+
+
+OK, thanks a lot for your response ... so let's close this bug now.
+