[Qemu-devel] [Bug 1795369] [NEW] Record/replay (icount rr) causes emulation hang or exit with error about missing events in log

[Artem Pisarenko]

Public bug reported:

Test case description:

Guest image is Linux, which just powers off after kernel boots (instead of 
proceeding to user-space /init or /sbin/init).
Base cmdline:
  qemu-system-x86_64 \
    -nodefaults -nographic -machine pc,accel=tcg -m 2048 -cpu qemu64 \
    -kernel bzImage -initrd rootfs -append 'nokaslr console=ttyS0 
rdinit=/init_poweroff' \
    -serial SERIAL_VALUE \
    -rtc clock=vm,base=2000-01-01T00:00:00 \
    -icount 1,sleep=off,rr=RR_VALUE,rrfile=icount_rr_capture.bin

Test 1.
When SERIAL_VALUE=none
Running with RR_VALUE=record completes successfully.
Running with RR_VALUE=replay doesn't completes. qemu process just eating ~100% 
cpu and memory usage doesn't grow after some moment. I don't see what happens 
because of problem no.2 (see below).

Test 2.
When SERIAL_VALUE=stdio
Running with RR_VALUE=record completes successfully.
Running with RR_VALUE=replay causes exit with error:
"qemu-system-x86_64: Missing character write event in the replay log"

Tests 3,4,5...
SERIAL_VALUE=stdio. Playing with "-rtc" clock and base suboptions, "-icount" 
sleep suboptions produces non-repeatable results.
In most cases running with RR_VALUE=record completes successfully (but may hang 
at very begining).
Running with RR_VALUE=replay with combinations of removing "-rtc base=..." and 
"-icount sleep=..." goes better, but at different places of boot process it may 
either hang (as in test 1) or exit with error (as in test 2).
When qemu "hangs", it may also happen differently: either it can be stopped by 
Ctrl-C, or have to be killed.


Guest image uploaded here: 
https://drive.google.com/open?id=1SHG4HyBdcPutc5Au4pyhN8z9w52et51A

QEMU built from master (commit 042938f46e1c477419d1931381fdadffaa49d45e) with:
<SRC_ROOT>/configure --prefix=<INSTALL_ROOT> --target-list=x86_64-softmmu 
--enable-debug --disable-pie --enable-tcg --disable-tcg-interpreter 
--enable-virtfs --disable-docs --disable-guest-agent --disable-modules 
--disable-gnutls --disable-nettle --disable-gcrypt --disable-sdl 
--disable-curses --disable-vnc --disable-vnc-sasl --disable-vnc-jpeg 
--disable-vnc-png --disable-cocoa --disable-xen --disable-xen-pci-passthrough 
--disable-brlapi --disable-curl --disable-fdt --disable-bluez --disable-kvm 
--disable-hax --disable-hvf --disable-whpx --disable-rdma --disable-vde 
--disable-netmap --disable-cap-ng --disable-spice --disable-rbd 
--disable-libiscsi --disable-libnfs --disable-smartcard --disable-libusb 
--disable-live-block-migration --disable-usb-redir --disable-glusterfs 
--disable-tpm --disable-libssh2 --disable-numa --disable-libxml2 
--disable-opengl --disable-virglrenderer --disable-qom-cast-debug 
--disable-tools --disable-vxhs --disable-crypto-afalg --disable-capstone 
--disable-replication --disable-xfsctl --disable-seccomp --disable-pvrdma 
--disable-libpmem

** Affects: qemu
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1795369

Title:
  Record/replay (icount rr) causes emulation hang or exit with error
  about missing events in log

Status in QEMU:
  New

Bug description:
  Test case description:

  Guest image is Linux, which just powers off after kernel boots (instead of 
proceeding to user-space /init or /sbin/init).
  Base cmdline:
    qemu-system-x86_64 \
      -nodefaults -nographic -machine pc,accel=tcg -m 2048 -cpu qemu64 \
      -kernel bzImage -initrd rootfs -append 'nokaslr console=ttyS0 
rdinit=/init_poweroff' \
      -serial SERIAL_VALUE \
      -rtc clock=vm,base=2000-01-01T00:00:00 \
      -icount 1,sleep=off,rr=RR_VALUE,rrfile=icount_rr_capture.bin

  Test 1.
  When SERIAL_VALUE=none
  Running with RR_VALUE=record completes successfully.
  Running with RR_VALUE=replay doesn't completes. qemu process just eating 
~100% cpu and memory usage doesn't grow after some moment. I don't see what 
happens because of problem no.2 (see below).

  Test 2.
  When SERIAL_VALUE=stdio
  Running with RR_VALUE=record completes successfully.
  Running with RR_VALUE=replay causes exit with error:
  "qemu-system-x86_64: Missing character write event in the replay log"

  Tests 3,4,5...
  SERIAL_VALUE=stdio. Playing with "-rtc" clock and base suboptions, "-icount" 
sleep suboptions produces non-repeatable results.
  In most cases running with RR_VALUE=record completes successfully (but may 
hang at very begining).
  Running with RR_VALUE=replay with combinations of removing "-rtc base=..." 
and "-icount sleep=..." goes better, but at different places of boot process it 
may either hang (as in test 1) or exit with error (as in test 2).
  When qemu "hangs", it may also happen differently: either it can be stopped 
by Ctrl-C, or have to be killed.

  
  Guest image uploaded here: 
https://drive.google.com/open?id=1SHG4HyBdcPutc5Au4pyhN8z9w52et51A

  QEMU built from master (commit 042938f46e1c477419d1931381fdadffaa49d45e) with:
  <SRC_ROOT>/configure --prefix=<INSTALL_ROOT> --target-list=x86_64-softmmu 
--enable-debug --disable-pie --enable-tcg --disable-tcg-interpreter 
--enable-virtfs --disable-docs --disable-guest-agent --disable-modules 
--disable-gnutls --disable-nettle --disable-gcrypt --disable-sdl 
--disable-curses --disable-vnc --disable-vnc-sasl --disable-vnc-jpeg 
--disable-vnc-png --disable-cocoa --disable-xen --disable-xen-pci-passthrough 
--disable-brlapi --disable-curl --disable-fdt --disable-bluez --disable-kvm 
--disable-hax --disable-hvf --disable-whpx --disable-rdma --disable-vde 
--disable-netmap --disable-cap-ng --disable-spice --disable-rbd 
--disable-libiscsi --disable-libnfs --disable-smartcard --disable-libusb 
--disable-live-block-migration --disable-usb-redir --disable-glusterfs 
--disable-tpm --disable-libssh2 --disable-numa --disable-libxml2 
--disable-opengl --disable-virglrenderer --disable-qom-cast-debug 
--disable-tools --disable-vxhs --disable-crypto-afalg --disable-capstone 
--disable-replication --disable-xfsctl --disable-seccomp --disable-pvrdma 
--disable-libpmem

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1795369/+subscriptions