cpu-exec: Add sleeping algorithm

The goal is to sleep qemu whenever the guest clock
is in advance compared to the host clock (we use
the monotonic clocks). The amount of time to sleep
is calculated in the execution loop in cpu_exec.

At first, we tried to approximate at each for loop the real time elapsed
while searching for a TB (generating or retrieving from cache) and
executing it. We would then approximate the virtual time corresponding
to the number of virtual instructions executed. The difference between
these 2 values would allow us to know if the guest is in advance or delayed.
However, the function used for measuring the real time
(qemu_clock_get_ns(QEMU_CLOCK_REALTIME)) proved to be very expensive.
We had an added overhead of 13% of the total run time.

Therefore, we modified the algorithm and only take into account the
difference between the 2 clocks at the begining of the cpu_exec function.
During the for loop we try to reduce the advance of the guest only by
computing the virtual time elapsed and sleeping if necessary. The overhead
is thus reduced to 3%. Even though this method still has a noticeable
overhead, it no longer is a bottleneck in trying to achieve a better
guest frequency for which the guest clock is faster than the host one.

As for the the alignement of the 2 clocks, with the first algorithm
the guest clock was oscillating between -1 and 1ms compared to the host clock.
Using the second algorithm we notice that the guest is 5ms behind the host, which
is still acceptable for our use case.

The tests where conducted using fio and stress. The host machine in an i5 CPU at
3.10GHz running Debian Jessie (kernel 3.12). The guest machine is an arm versatile-pb
built with buildroot.

Currently, on our test machine, the lowest icount we can achieve that is suitable for
aligning the 2 clocks is 6. However, we observe that the IO tests (using fio) are
slower than the cpu tests (using stress).

Signed-off-by: Sebastian Tanase <sebastian.tanase@openwide.fr>
Tested-by: Camille Bégué <camille.begue@openwide.fr>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
3 files changed
tree: 09c74582ec42dfebee60217a01665e784eb20747
  1. audio/
  2. backends/
  3. block/
  4. bsd-user/
  5. default-configs/
  6. disas/
  7. docs/
  8. fpu/
  9. fsdev/
  10. gdb-xml/
  11. hw/
  12. include/
  13. libcacard/
  14. libdecnumber/
  15. linux-headers/
  16. linux-user/
  17. net/
  18. pc-bios/
  19. po/
  20. qapi/
  21. qga/
  22. qobject/
  23. qom/
  24. roms/
  25. scripts/
  26. slirp/
  27. stubs/
  28. sysconfigs/
  29. target-alpha/
  30. target-arm/
  31. target-cris/
  32. target-i386/
  33. target-lm32/
  34. target-m68k/
  35. target-microblaze/
  36. target-mips/
  37. target-moxie/
  38. target-openrisc/
  39. target-ppc/
  40. target-s390x/
  41. target-sh4/
  42. target-sparc/
  43. target-unicore32/
  44. target-xtensa/
  45. tcg/
  46. tests/
  47. trace/
  48. ui/
  49. util/
  50. .exrc
  51. .gitignore
  52. .gitmodules
  53. .mailmap
  54. .travis.yml
  55. aio-posix.c
  56. aio-win32.c
  57. arch_init.c
  58. async.c
  59. balloon.c
  60. block-migration.c
  61. block.c
  62. blockdev-nbd.c
  63. blockdev.c
  64. blockjob.c
  65. bt-host.c
  66. bt-vhci.c
  67. Changelog
  68. CODING_STYLE
  69. configure
  70. COPYING
  71. COPYING.LIB
  72. coroutine-gthread.c
  73. coroutine-sigaltstack.c
  74. coroutine-ucontext.c
  75. coroutine-win32.c
  76. cpu-exec.c
  77. cpus.c
  78. cputlb.c
  79. device-hotplug.c
  80. device_tree.c
  81. disas.c
  82. dma-helpers.c
  83. dump.c
  84. exec.c
  85. gdbstub.c
  86. HACKING
  87. hmp-commands.hx
  88. hmp.c
  89. hmp.h
  90. iohandler.c
  91. ioport.c
  92. iothread.c
  93. kvm-all.c
  94. kvm-stub.c
  95. LICENSE
  96. main-loop.c
  97. MAINTAINERS
  98. Makefile
  99. Makefile.objs
  100. Makefile.target
  101. memory.c
  102. memory_mapping.c
  103. migration-exec.c
  104. migration-fd.c
  105. migration-rdma.c
  106. migration-tcp.c
  107. migration-unix.c
  108. migration.c
  109. module-common.c
  110. monitor.c
  111. nbd.c
  112. numa.c
  113. os-posix.c
  114. os-win32.c
  115. page_cache.c
  116. qapi-schema.json
  117. qdev-monitor.c
  118. qdict-test-data.txt
  119. qemu-bridge-helper.c
  120. qemu-char.c
  121. qemu-coroutine-io.c
  122. qemu-coroutine-lock.c
  123. qemu-coroutine-sleep.c
  124. qemu-coroutine.c
  125. qemu-doc.texi
  126. qemu-file.c
  127. qemu-img-cmds.hx
  128. qemu-img.c
  129. qemu-img.texi
  130. qemu-io-cmds.c
  131. qemu-io.c
  132. qemu-log.c
  133. qemu-nbd.c
  134. qemu-nbd.texi
  135. qemu-options-wrapper.h
  136. qemu-options.h
  137. qemu-options.hx
  138. qemu-seccomp.c
  139. qemu-tech.texi
  140. qemu-timer.c
  141. qemu.nsi
  142. qemu.sasl
  143. qmp-commands.hx
  144. qmp.c
  145. qtest.c
  146. README
  147. rules.mak
  148. savevm.c
  149. softmmu_template.h
  150. spice-qemu-char.c
  151. tcg-runtime.c
  152. tci.c
  153. thread-pool.c
  154. thunk.c
  155. tpm.c
  156. trace-events
  157. translate-all.c
  158. translate-all.h
  159. user-exec.c
  160. VERSION
  161. version.rc
  162. vl.c
  163. vmstate.c
  164. xbzrle.c
  165. xen-common-stub.c
  166. xen-common.c
  167. xen-hvm-stub.c
  168. xen-hvm.c
  169. xen-mapcache.c