| Message ID | 20170123232738.20796-1-ehabkost@redhat.com |
|---|---|
| State | New |
| Headers | show |
On 23 January 2017 at 23:27, Eduardo Habkost <ehabkost@redhat.com> wrote: > The following changes since commit 3879284d6517dc22529395bdb259f4183b589127: > > Merge remote-tracking branch 'remotes/berrange/tags/pull-qio-2017-01-23-2' into staging (2017-01-23 15:59:09 +0000) > > are available in the git repository at: > > git://github.com/ehabkost/qemu.git tags/x86-and-machine-pull-request > > for you to fetch changes up to d99569d9d8562c480e0befab601756b0b7b5d0e0: > > kvm: Allow invtsc migration if tsc-khz is set explicitly (2017-01-23 21:26:27 -0200) > > ---------------------------------------------------------------- > x86, machine, numa queue (2017-01-23) > > ---------------------------------------------------------------- I get a failure in the docs on OSX: GEN qemu-doc.txt CC qapi/qapi-visit-core.o /Users/pm215/src/qemu-for-merges/qemu-doc.texi:7: warning: unrecognized encoding name `UTF-8'. GEN trace/generated-tracers.c /Users/pm215/src/qemu-for-merges/qemu-doc.texi:1419: Cross reference to nonexistent node `bt-hcis' (perhaps incorrect sectioning?). makeinfo: Removing output file `qemu-doc.txt' due to errors; use --force to preserve. This is the same error that was happening on the Travis OSX builds but not on my personal OSX machine, and leaves me suspicious that the qemu-doc.txt stuff is broken but also not always being rebuilt correctly (since it's a bit hit-and-miss whether particular things hit this.) thanks -- PMM
On 24 January 2017 at 13:24, Peter Maydell <peter.maydell@linaro.org> wrote: > I get a failure in the docs on OSX: > GEN qemu-doc.txt > CC qapi/qapi-visit-core.o > /Users/pm215/src/qemu-for-merges/qemu-doc.texi:7: warning: > unrecognized encoding name `UTF-8'. > GEN trace/generated-tracers.c > /Users/pm215/src/qemu-for-merges/qemu-doc.texi:1419: Cross reference > to nonexistent node `bt-hcis' (perhaps incorrect sectioning?). > makeinfo: Removing output file `qemu-doc.txt' due to errors; use > --force to preserve. > > This is the same error that was happening on the Travis OSX > builds but not on my personal OSX machine, and leaves me > suspicious that the qemu-doc.txt stuff is broken but also > not always being rebuilt correctly (since it's a bit > hit-and-miss whether particular things hit this.) The cause of this problem I think is that we added a new output file qemu-doc.txt, but we forgot to add it to this dependency rule in Makefile: qemu-doc.html qemu-doc.info qemu-doc.pdf: \ qemu-img.texi qemu-nbd.texi qemu-options.texi qemu-option-trace.texi \ qemu-monitor.texi qemu-img-cmds.texi qemu-ga.texi \ qemu-monitor-info.texi So if you do a make with -j8 then make may decide to parallelize the building of the qemu-doc.txt with the creation of qemu-options.texi, and then makeinfo can read a half-completed qemu-options.texi. qemu-doc.txt also seems to have been not added to the rm -f commands for the distclean target. Side question: why do we have a rule to create the qemu-doc.pdf but not actually install it with install-doc ? thanks -- PMM
Peter Maydell <peter.maydell@linaro.org> writes: > On 24 January 2017 at 13:24, Peter Maydell <peter.maydell@linaro.org> wrote: >> I get a failure in the docs on OSX: >> GEN qemu-doc.txt >> CC qapi/qapi-visit-core.o >> /Users/pm215/src/qemu-for-merges/qemu-doc.texi:7: warning: >> unrecognized encoding name `UTF-8'. >> GEN trace/generated-tracers.c >> /Users/pm215/src/qemu-for-merges/qemu-doc.texi:1419: Cross reference >> to nonexistent node `bt-hcis' (perhaps incorrect sectioning?). >> makeinfo: Removing output file `qemu-doc.txt' due to errors; use >> --force to preserve. >> >> This is the same error that was happening on the Travis OSX >> builds but not on my personal OSX machine, and leaves me >> suspicious that the qemu-doc.txt stuff is broken but also >> not always being rebuilt correctly (since it's a bit >> hit-and-miss whether particular things hit this.) > > The cause of this problem I think is that we added a new > output file qemu-doc.txt, but we forgot to add it to this > dependency rule in Makefile: > > qemu-doc.html qemu-doc.info qemu-doc.pdf: \ > qemu-img.texi qemu-nbd.texi qemu-options.texi qemu-option-trace.texi \ > qemu-monitor.texi qemu-img-cmds.texi qemu-ga.texi \ > qemu-monitor-info.texi > > So if you do a make with -j8 then make may decide to parallelize > the building of the qemu-doc.txt with the creation of qemu-options.texi, > and then makeinfo can read a half-completed qemu-options.texi. Good catch! Now, why didn't I see that myself? %-} > qemu-doc.txt also seems to have been not added to the rm -f > commands for the distclean target. > > Side question: why do we have a rule to create the qemu-doc.pdf > but not actually install it with install-doc ? I don't know. Has been that way since the pdf target was added in commit 20cc999.
On 23 January 2017 at 23:27, Eduardo Habkost <ehabkost@redhat.com> wrote: > The following changes since commit 3879284d6517dc22529395bdb259f4183b589127: > > Merge remote-tracking branch 'remotes/berrange/tags/pull-qio-2017-01-23-2' into staging (2017-01-23 15:59:09 +0000) > > are available in the git repository at: > > git://github.com/ehabkost/qemu.git tags/x86-and-machine-pull-request > > for you to fetch changes up to d99569d9d8562c480e0befab601756b0b7b5d0e0: > > kvm: Allow invtsc migration if tsc-khz is set explicitly (2017-01-23 21:26:27 -0200) > > ---------------------------------------------------------------- > x86, machine, numa queue (2017-01-23) > > ---------------------------------------------------------------- Applied, thanks (since we've tracked down the makeinfo issue). -- PMM
The following changes since commit 3879284d6517dc22529395bdb259f4183b589127: Merge remote-tracking branch 'remotes/berrange/tags/pull-qio-2017-01-23-2' into staging (2017-01-23 15:59:09 +0000) are available in the git repository at: git://github.com/ehabkost/qemu.git tags/x86-and-machine-pull-request for you to fetch changes up to d99569d9d8562c480e0befab601756b0b7b5d0e0: kvm: Allow invtsc migration if tsc-khz is set explicitly (2017-01-23 21:26:27 -0200) ---------------------------------------------------------------- x86, machine, numa queue (2017-01-23) ---------------------------------------------------------------- Dou Liyang (1): vl: Ensure the numa_post_machine_init func in the appropriate location Eduardo Habkost (6): i386: Remove AMD feature flag aliases from Opteron models i386: Return migration-safe field on query-cpu-definitions arch_init: Remove unnecessary default_config_files table qemu-options: Rename variables on the -numa "cpus" option kvm: Simplify invtsc check kvm: Allow invtsc migration if tsc-khz is set explicitly He Chen (1): x86: add AVX512_VPOPCNTDQ features Igor Mammedov (3): pc: cleanup: move smbios_set_cpuid() into pc_build_smbios() pc: don't return cpu pointer from pc_new_cpu() as it's not needed anymore machine: Make possible_cpu_arch_ids() return const pointer Thomas Huth (2): MAINTAINERS: Add an entry for hw/core/null-machine.c hw/core/null-machine: Add the possibility to instantiate a CPU and RAM include/hw/acpi/acpi_dev_interface.h | 2 +- include/hw/boards.h | 2 +- include/hw/i386/pc.h | 2 +- include/qemu/config-file.h | 4 --- target/i386/cpu-qom.h | 2 ++ target/i386/cpu.h | 1 + arch_init.c | 27 ------------------- hw/acpi/cpu.c | 6 ++--- hw/acpi/cpu_hotplug.c | 4 +-- hw/core/null-machine.c | 27 +++++++++++++++++-- hw/i386/acpi-build.c | 8 +++--- hw/i386/pc.c | 51 ++++++++++++++---------------------- stubs/pc_madt_cpu_entry.c | 2 +- target/i386/cpu.c | 51 +++++++++++------------------------- target/i386/kvm.c | 20 +++++++------- vl.c | 23 +++++++++++----- MAINTAINERS | 1 + hw/core/Makefile.objs | 2 +- qemu-options.hx | 8 +++--- 19 files changed, 108 insertions(+), 135 deletions(-)