Move target- and device specific code from monitor
v2 cover Move target- and device specific code from monitor
The monivation of this set is simple. Recently we have proposed patch to monitor.c with specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code in monitor.c" This patchset is the first step to move arch specific code from monitor.c targets.
So, monitor.c already contains a lot of generic code, as well as the target specifics code and eventually monitor.c volume will only grow. This trend leads to a variety of fouling code ifdeffery(and combinations thereof), poor readability, and entanglement of architecture of the project. If someone wants to improve processing logic commands at the monitor, it isn't necessarily must differentiate amongst the implementation of some ARM or x86_64 specific commands, because the project already has separation of target specific code on directories.
The presented solution is not the best, but it is quite simple (PATCH doesn't add more code!) and decides the above mentioned issue. Subsequently it will not prevent the introduction of more advanced mechanism that can more effectively resolve the issue.
There is a issue with the placement of code for multiple architectures (isn't for everyone), but this code is very small. This patch is a step towards solving the issue associated with maintaining the purity of the code and structure of the project, which solves not all, but doing a little better than it is.
Changes from v1:
- ported to new head
PATCH v2 1/3 hmp-commands-info: move info_cmds content out of monitor.c
From: Pavel Butsykin <[email protected]>
For moving target- and device-specific code from monitor.c, to beginning we move info_cmds content to hmp-commands-info.hx
--- Makefile.target | 5 +- hmp-commands-info.hx | 711 +++++++++++++++++++++++++++++++++++++++++++++++++++ monitor.c | 373 +-------------------------- 3 files changed, 717 insertions(+), 372 deletions(-) create mode 100644 hmp-commands-info.hx
Makefile.target +4 -1
@@ -151,7 +151,7 @@ else obj-y += hw/$(TARGET_BASE_ARCH)/ endif -GENERATED_HEADERS += hmp-commands.h qmp-commands-old.h +GENERATED_HEADERS += hmp-commands.h hmp-commands-info.h qmp-commands-old.h endif # CONFIG_SOFTMMU
@@ -193,6 +193,9 @@ gdbstub-xml.c: $(TARGET_XML_FILES) $(SRC_PATH)/scripts/feature_to_c.sh hmp-commands.h: $(SRC_PATH)/hmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -h < $< > $@," GEN $(TARGET_DIR)$@") +hmp-commands-info.h: $(SRC_PATH)/hmp-commands-info.hx + $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -h < $< > $@," GEN $(TARGET_DIR)$@") + qmp-commands-old.h: $(SRC_PATH)/qmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -h < $< > $@," GEN $(TARGET_DIR)$@")
hmp-commands-info.hx new file +711 -0
711 lines not on this page · raw message
monitor.c +2 -371
379 lines not on this page · raw message
Signature · 1 line
2.1.4
Re: [Qemu-devel] [PATCH 1/3] hmp-commands-info: move info_cmds content out of monitor.c
Quoted 11 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
For moving target- and device-specific code from
monitor.c, to beginning we moveinfo_cmdscontent to hmp-commands-info.hxSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>---
We should be able to pull some of the #ifdefs out of this a bit better, but easier to do that as a secondary stop.
thanks
-- PMM
PATCH v2 2/3 monitor: remove target-specific code from monitor.c
In reply to Denis V. LunevFrom: Pavel Butsykin <[email protected]>
Move target-specific code out of /monitor.c to /target-*/monitor.c, this will avoid code cluttering and using random ifdeffery. The solution is quite simple, but solves the issue of the separation of target-specific code from monitor
--- include/monitor/monitor-common.h | 43 ++ monitor.c | 854 +-------------------------------------- target-i386/Makefile.objs | 2 +- target-i386/monitor.c | 489 ++++++++++++++++++++++ target-ppc/Makefile.objs | 2 +- target-ppc/monitor.c | 250 ++++++++++++ target-sh4/Makefile.objs | 1 + target-sh4/monitor.c | 52 +++ target-sparc/Makefile.objs | 2 +- target-sparc/monitor.c | 153 +++++++ target-xtensa/Makefile.objs | 1 + target-xtensa/monitor.c | 34 ++ 12 files changed, 1032 insertions(+), 851 deletions(-) create mode 100644 include/monitor/monitor-common.h create mode 100644 target-i386/monitor.c create mode 100644 target-ppc/monitor.c create mode 100644 target-sh4/monitor.c create mode 100644 target-sparc/monitor.c create mode 100644 target-xtensa/monitor.c
include/monitor/monitor-common.h new file +43 -0
@@ -0,0 +1,43 @@ +/* +* QEMU monitor +* +* Copyright (c) 2003-2004 Fabrice Bellard +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to deal +* in the Software without restriction, including without limitation the rights +* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +* copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +* THE SOFTWARE. +*/ +#ifndef MONITOR_COMMON_H +#define MONITOR_COMMON_H + +#define MD_TLONG 0 +#define MD_I32 1 + +typedef struct MonitorDef { +const char *name; +int offset; +target_long (*get_value)(const struct MonitorDef *md,int val); +int type; +}MonitorDef ; + +CPUArchState *mon_get_cpu_env (void ); + +void hmp_info_mem (Monitor *mon,const QDict *qdict); +void hmp_info_tlb (Monitor *mon,const QDict *qdict); +void hmp_mce (Monitor *mon,const QDict *qdict); + +#endif /* MONITOR_COMMON */
monitor.c +6 -848
886 lines not on this page · raw message
target-i386/Makefile.objs +1 -1
@@ -2,6 +2,6 @@ obj-y += translate.o helper.o cpu.o obj-y += excp_helper.o fpu_helper.o cc_helper.o int_helper.o svm_helper.o obj-y += smm_helper.o misc_helper.o mem_helper.o seg_helper.o obj-y += gdbstub.o -obj-$(CONFIG_SOFTMMU) += machine.o arch_memory_mapping.o arch_dump.o +obj-$(CONFIG_SOFTMMU) += machine.o arch_memory_mapping.o arch_dump.o monitor.o obj-$(CONFIG_KVM) += kvm.o obj-$(call lnot,$(CONFIG_KVM)) += kvm-stub.o
target-i386/monitor.c new file +489 -0
489 lines not on this page · raw message
target-ppc/Makefile.objs +1 -1
@@ -1,7 +1,7 @@ obj-y += cpu-models.o obj-y += translate.o ifeq ($(CONFIG_SOFTMMU),y) -obj-y += machine.o mmu_helper.o mmu-hash32.o +obj-y += machine.o mmu_helper.o mmu-hash32.o monitor.o obj-$(TARGET_PPC64) += mmu-hash64.o arch_dump.o endif obj-$(CONFIG_KVM) += kvm.o kvm_ppc.o
target-ppc/monitor.c new file +250 -0
250 lines not on this page · raw message
target-sh4/Makefile.objs +1 -0
@@ -1,2 +1,3 @@ obj-y += translate.o op_helper.o helper.o cpu.o +obj-$(CONFIG_SOFTMMU) += monitor.o obj-y += gdbstub.o
target-sh4/monitor.c new file +52 -0
52 lines not on this page · raw message
target-sparc/Makefile.objs +1 -1
@@ -1,4 +1,4 @@ -obj-$(CONFIG_SOFTMMU) += machine.o +obj-$(CONFIG_SOFTMMU) += machine.o monitor.o obj-y += translate.o helper.o cpu.o obj-y += fop_helper.o cc_helper.o win_helper.o mmu_helper.o ldst_helper.o obj-$(TARGET_SPARC) += int32_helper.o
target-sparc/monitor.c new file +153 -0
153 lines not on this page · raw message
target-xtensa/Makefile.objs +1 -0
@@ -2,5 +2,6 @@ obj-y += xtensa-semi.o obj-y += core-dc232b.o obj-y += core-dc233c.o obj-y += core-fsf.o +obj-$(CONFIG_SOFTMMU) += monitor.o obj-y += translate.o op_helper.o helper.o cpu.o obj-y += gdbstub.o
target-xtensa/monitor.c new file +34 -0
@@ -0,0 +1,34 @@ +/* +* QEMU monitor +* +* Copyright (c) 2003-2004 Fabrice Bellard +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to deal +* in the Software without restriction, including without limitation the rights +* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +* copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +* THE SOFTWARE. +*/ +#include "cpu.h" +#include "monitor/monitor.h" +#include "monitor/monitor-common.h" +#include "hmp.h" + +void hmp_info_tlb (Monitor *mon,const QDict *qdict) +{ +CPUArchState *env1 =mon_get_cpu_env (); + +dump_mmu ((FILE *)mon, (fprintf_function )monitor_printf, env1); +}
Signature · 1 line
2.1.4
Re: [Qemu-devel] [PATCH 2/3] monitor: remove target-specific code from monitor.c
Quoted 33 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
Move target-specific code out of /monitor.c to /target-*/monitor.c, this will avoid code cluttering and using random ifdeffery. The solution is quite simple, but solves the issue of the separation of target-specific code from monitor
Signed-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>--- include/monitor/monitor-common.h | 43 ++ monitor.c | 854 +-------------------------------------- target-i386/Makefile.objs | 2 +- target-i386/monitor.c | 489 ++++++++++++++++++++++ target-ppc/Makefile.objs | 2 +- target-ppc/monitor.c | 250 ++++++++++++ target-sh4/Makefile.objs | 1 + target-sh4/monitor.c | 52 +++ target-sparc/Makefile.objs | 2 +- target-sparc/monitor.c | 153 +++++++ target-xtensa/Makefile.objs | 1 + target-xtensa/monitor.c | 34 ++ 12 files changed, 1032 insertions(+), 851 deletions(-) create mode 100644 include/monitor/monitor-common.h create mode 100644 target-i386/monitor.c create mode 100644 target-ppc/monitor.c create mode 100644 target-sh4/monitor.c create mode 100644 target-sparc/monitor.c create mode 100644 target-xtensa/monitor.c
monitor.c static const mon_cmd_t qmp_cmds[] = { · line 2076
Quoted 5 lines
monitor.c +4
-0+#if defined(TARGET_SPARC ) || defined(TARGET_PPC ) || defined(TARGET_I386 ) +extern const MonitorDef monitor_defs[]; +#else +const MonitorDef monitor_defs[] = {};#endif
So, rather than having to have a list of which targets provide a monitor_defs[], I would suggest that we make the API implemented by the target be a function, like:
const MonitorDef *target_monitor_defs(void);
(which just returns a pointer to a static const array in the target-*/monitor.c file). Then you can add a file stubs/target-monitor-defs.c which provides the "stub" version of this function (just returns a pointer to the no-commands array). The link process will arrange that the stub version is pulled in for any target that doesn't provide its own implementation of the function.
Other than that, I suspect we can improve the separation out of target-specific things, but this is a good improvement and it'll be easier to do the rest as incremental fixes on top of this later.
thanks
-- PMM
Re: [Qemu-devel] [PATCH 2/3] monitor: remove target-specific code from monitor.c
Quoted 50 lines · Peter Maydell
Quoted 32 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
Move target-specific code out of /monitor.c to /target-*/monitor.c, this will avoid code cluttering and using random ifdeffery. The solution is quite simple, but solves the issue of the separation of target-specific code from monitor
Signed-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>--- include/monitor/monitor-common.h | 43 ++ monitor.c | 854 +-------------------------------------- target-i386/Makefile.objs | 2 +- target-i386/monitor.c | 489 ++++++++++++++++++++++ target-ppc/Makefile.objs | 2 +- target-ppc/monitor.c | 250 ++++++++++++ target-sh4/Makefile.objs | 1 + target-sh4/monitor.c | 52 +++ target-sparc/Makefile.objs | 2 +- target-sparc/monitor.c | 153 +++++++ target-xtensa/Makefile.objs | 1 + target-xtensa/monitor.c | 34 ++ 12 files changed, 1032 insertions(+), 851 deletions(-)create mode 100644
include/monitor/monitor-common.hcreate mode 100644target-i386/monitor.ccreate mode 100644target-ppc/monitor.ccreate mode 100644target-sh4/monitor.ccreate mode 100644target-sparc/monitor.ccreate mode 100644target-xtensa/monitor.cmonitor.c static const mon_cmd_t qmp_cmds[] = { · line 2076
Quoted 5 lines
monitor.c +4
-0+#if defined(TARGET_SPARC ) || defined(TARGET_PPC ) || defined(TARGET_I386 ) +extern const MonitorDef monitor_defs[]; +#else +const MonitorDef monitor_defs[] = {};#endif So, rather than having to have a list of which targets provide a monitor_defs[], I would suggest that we make the API implemented by the target be a function, like:
const MonitorDef *target_monitor_defs(void);(which just returns a pointer to a static const array in the
target-*/monitor.cfile). Then you can add a filestubs/target-monitor-defs.cwhich provides the "stub" version of this function (just returns a pointer to the no-commands array). The link process will arrange that the stub version is pulled in for any target that doesn't provide its own implementation of the function.
Quoted 7 lines
Other than that, I suspect we can improve the separation out of target-specific things, but this is a good improvement and it'll be easier to do the rest as incremental fixes on top of this later.
thanks
-- PMM
ok, sounds good to me
Thank you for the suggestion
Re: [Qemu-devel] [PATCH 2/3] monitor: remove target-specific code from monitor.c
In reply to Peter MaydellQuoted 50 lines · Peter Maydell
Quoted 32 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
Move target-specific code out of /monitor.c to /target-*/monitor.c, this will avoid code cluttering and using random ifdeffery. The solution is quite simple, but solves the issue of the separation of target-specific code from monitor
Signed-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>--- include/monitor/monitor-common.h | 43 ++ monitor.c | 854 +-------------------------------------- target-i386/Makefile.objs | 2 +- target-i386/monitor.c | 489 ++++++++++++++++++++++ target-ppc/Makefile.objs | 2 +- target-ppc/monitor.c | 250 ++++++++++++ target-sh4/Makefile.objs | 1 + target-sh4/monitor.c | 52 +++ target-sparc/Makefile.objs | 2 +- target-sparc/monitor.c | 153 +++++++ target-xtensa/Makefile.objs | 1 + target-xtensa/monitor.c | 34 ++ 12 files changed, 1032 insertions(+), 851 deletions(-)create mode 100644
include/monitor/monitor-common.hcreate mode 100644target-i386/monitor.ccreate mode 100644target-ppc/monitor.ccreate mode 100644target-sh4/monitor.ccreate mode 100644target-sparc/monitor.ccreate mode 100644target-xtensa/monitor.cmonitor.c static const mon_cmd_t qmp_cmds[] = { · line 2076
Quoted 5 lines
monitor.c +4
-0+#if defined(TARGET_SPARC ) || defined(TARGET_PPC ) || defined(TARGET_I386 ) +extern const MonitorDef monitor_defs[]; +#else +const MonitorDef monitor_defs[] = {};#endif So, rather than having to have a list of which targets provide a monitor_defs[], I would suggest that we make the API implemented by the target be a function, like:
const MonitorDef *target_monitor_defs(void);(which just returns a pointer to a static const array in the
target-*/monitor.cfile). Then you can add a filestubs/target-monitor-defs.cwhich provides the "stub" version of this function (just returns a pointer to the no-commands array). The link process will arrange that the stub version is pulled in for any target that doesn't provide its own implementation of the function.
Quoted 7 lines
Other than that, I suspect we can improve the separation out of target-specific things, but this is a good improvement and it'll be easier to do the rest as incremental fixes on top of this later.
thanks
-- PMM
Yes, this is a good way if we make the interface: const MonitorDef *target_monitor_defs(void); But we can't include the 'monitor/monitor-common.h' to stubs/target-monitor-defs.c, because there is a dependency with a target-specific headers( such as cpu.h:CPUArchState, cpu-defs.h:target_long). Make a copy of the struct MonitorDef not a good way because we can miss the change of copied MonitorDef in stubs/target-monitor-defs.c and this will result in an bug. Can this be solved somehow else?
Re: [Qemu-devel] [PATCH 2/3] monitor: remove target-specific code from monitor.c
Quoted 10 lines · Pavel Butsykin
Yes, this is a good way if we make the interface: const MonitorDef
*target_monitor_defs(void);
But we can't include the 'monitor/monitor-common.h' tostubs/target-monitor-defs.c, because
there is a dependency with a target-specificheaders( such as cpu.h:CPUArchState, cpu-defs.h:target_long).
Make a copy of the struct MonitorDef not a good way because we can miss the change of copied MonitorDef
instubs/target-monitor-defs.cand this will result in an bug. Can this be solved somehow else?
That target_long is awkward. I suggest making the stub version just return NULL rather than a pointer to an empty list (you'll need to adjust the calling code slightly to handle the NULL return). Then the stub code only needs to be able to see the 'typedef struct MonitorDef MonitorDef' (which you can put in qemu-typedefs.h).
-- PMM
Re: [Qemu-devel] [PATCH 2/3] monitor: remove target-specific code from monitor.c
Quoted 11 lines · Peter Maydell
Quoted 10 lines · Pavel Butsykin
Yes, this is a good way if we make the interface: const MonitorDef
*target_monitor_defs(void);
But we can't include the 'monitor/monitor-common.h' tostubs/target-monitor-defs.c, because
there is a dependency with a target-specificheaders( such as cpu.h:CPUArchState, cpu-defs.h:target_long).
Make a copy of the struct MonitorDef not a good way because we can miss the change of copied MonitorDef
instubs/target-monitor-defs.cand this will result in an bug. Can this be solved somehow else?
Quoted 8 lines
That
target_longis awkward. I suggest making the stub version just return NULL rather than a pointer to an empty list (you'll need to adjust the calling code slightly to handle the NULL return). Then the stub code only needs to be able to see the 'typedef struct MonitorDef MonitorDef' (which you can put inqemu-typedefs.h).-- PMM
Thank you for the suggestion, this seems a good way out.
PATCH v2 3/3 monitor: added generation of documentation for hmp-commands-info.hx
In reply to Denis V. LunevFrom: Pavel Butsykin <[email protected]>
It will be easier if you need to add info-commands to edit only hmp-commands-info.hx, before this had to edit monitor.c and hmp-commands.hx
--- .gitignore | 1 + Makefile | 9 ++-- hmp-commands-info.hx | 4 ++ hmp-commands.hx | 120 --------------------------------------------------- qemu-doc.texi | 2 + 5 files changed, 13 insertions(+), 123 deletions(-)
.gitignore +1 -0
@@ -49,6 +49,7 @@ /qemu-ga /qemu-bridge-helper /qemu-monitor.texi +/qemu-monitor-info.texi /qmp-commands.txt /vscclient /fsdev/virtfs-proxy-helper
Makefile +6 -3
@@ -344,7 +344,7 @@ qemu-%.tar.bz2: $(SRC_PATH)/scripts/make-release "$(SRC_PATH)" "$(patsubst qemu-%.tar.bz2,%,$@)" distclean: clean - rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi + rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi qemu-monitor-info.texi rm -f config-all-devices.mak config-all-disas.mak config.status rm -f po/*.mo tests/qemu-iotests/common.env rm -f roms/seabios/config.mak roms/vgabios/config.mak
@@ -508,13 +508,16 @@ qemu-options.texi: $(SRC_PATH)/qemu-options.hx qemu-monitor.texi: $(SRC_PATH)/hmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") +qemu-monitor-info.texi: $(SRC_PATH)/hmp-commands-info.hx + $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") + qmp-commands.txt: $(SRC_PATH)/qmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -q < $< > $@," GEN $@") qemu-img-cmds.texi: $(SRC_PATH)/qemu-img-cmds.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") -qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi +qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi qemu-monitor-info.texi $(call quiet-command, \ perl -Ww -- $(SRC_PATH)/scripts/texi2pod.pl $< qemu.pod && \ $(POD2MAN) --section=1 --center=" " --release=" " qemu.pod > $@, \
@@ -551,7 +554,7 @@ pdf: qemu-doc.pdf qemu-tech.pdf qemu-doc.dvi qemu-doc.html qemu-doc.info qemu-doc.pdf: \ qemu-img.texi qemu-nbd.texi qemu-options.texi \ - qemu-monitor.texi qemu-img-cmds.texi + qemu-monitor.texi qemu-monitor-info.texi qemu-img-cmds.texi ifdef CONFIG_WIN32
hmp-commands-info.hx +4 -0
@@ -6,6 +6,9 @@ HXCOMM monitor info commands HXCOMM HXCOMM can be used for comments, discarded from both texi and C STEXI +@item info @var{subcommand} +@findex info +Show various information about the system state. @table @option ETEXI
@@ -708,4 +711,5 @@ ETEXI STEXI @end table +@end table ETEXI
hmp-commands.hx +0 -120
123 lines not on this page · raw message
qemu-doc.texi +2 -0
@@ -391,6 +391,8 @@ The following commands are available: @include qemu-monitor.texi +@include qemu-monitor-info.texi + @subsection Integer expressions The monitor understands integers expressions for every integer
Signature · 1 line
2.1.4
Re: [Qemu-devel] [PATCH 3/3] monitor: added generation of documentation for hmp-commands-info.hx
Quoted 86 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
It will be easier if you need to add info-commands to edit only hmp-commands-info.hx, before this had to edit
monitor.cand hmp-commands.hxSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>--- .gitignore | 1 + Makefile | 9 ++-- hmp-commands-info.hx | 4 ++ hmp-commands.hx | 120 --------------------------------------------------- qemu-doc.texi | 2 + 5 files changed, 13 insertions(+), 123 deletions(-).gitignore +1
-0@@ -49,6 +49,7 @@ /qemu-ga /qemu-bridge-helper /qemu-monitor.texi +/qemu-monitor-info.texi /qmp-commands.txt /vscclient /fsdev/virtfs-proxy-helperMakefile +6
-3@@ -344,7 +344,7 @@ qemu-%.tar.bz2: $(SRC_PATH)/scripts/make-release "$(SRC_PATH)" "$(patsubst qemu-%.tar.bz2,%,$@)" distclean: clean - rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi + rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi qemu-monitor-info.texi rm -f config-all-devices.mak config-all-disas.mak config.status rm -f po/*.mo tests/qemu-iotests/common.env rm -f roms/seabios/config.mak roms/vgabios/config.mak@@ -508,13 +508,16 @@ qemu-options.texi: $(SRC_PATH)/qemu-options.hx qemu-monitor.texi: $(SRC_PATH)/hmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") +qemu-monitor-info.texi: $(SRC_PATH)/hmp-commands-info.hx + $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") + qmp-commands.txt: $(SRC_PATH)/qmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -q < $< > $@," GEN $@") qemu-img-cmds.texi: $(SRC_PATH)/qemu-img-cmds.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") -qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi +qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi qemu-monitor-info.texi $(call quiet-command, \ perl -Ww -- $(SRC_PATH)/scripts/texi2pod.pl $< qemu.pod && \ $(POD2MAN) --section=1 --center=" " --release=" " qemu.pod > $@, \@@ -551,7 +554,7 @@ pdf: qemu-doc.pdf qemu-tech.pdf qemu-doc.dvi qemu-doc.html qemu-doc.info qemu-doc.pdf: \ qemu-img.texi qemu-nbd.texi qemu-options.texi \ - qemu-monitor.texi qemu-img-cmds.texi + qemu-monitor.texi qemu-monitor-info.texi qemu-img-cmds.texi ifdef CONFIG_WIN32hmp-commands-info.hx +3
-0@@ -6,6 +6,9 @@ HXCOMM monitor info commands HXCOMM HXCOMM can be used for comments, discarded from both texi and C STEXI +@item info @var{subcommand} +@findex info +Show various information about the system state. @table @option ETEXI
hmp-commands-info.hx ETEXI · line 715
Quoted 5 lines
hmp-commands-info.hx +1
-0@@ -708,4 +711,5 @@ ETEXI STEXI @end table +@end table ETEXI
Where does this extra "@end table" come from?
thanks
-- PMM
Re: [Qemu-devel] [PATCH 3/3] monitor: added generation of documentation for hmp-commands-info.hx
Quoted 88 lines · Peter Maydell
Quoted 87 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
It will be easier if you need to add info-commands to edit only hmp-commands-info.hx, before this had to edit
monitor.cand hmp-commands.hxSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>--- .gitignore | 1 + Makefile | 9 ++-- hmp-commands-info.hx | 4 ++ hmp-commands.hx | 120 --------------------------------------------------- qemu-doc.texi | 2 + 5 files changed, 13 insertions(+), 123 deletions(-).gitignore +1
-0@@ -49,6 +49,7 @@ /qemu-ga /qemu-bridge-helper /qemu-monitor.texi +/qemu-monitor-info.texi /qmp-commands.txt /vscclient /fsdev/virtfs-proxy-helperMakefile +6
-3@@ -344,7 +344,7 @@ qemu-%.tar.bz2: $(SRC_PATH)/scripts/make-release "$(SRC_PATH)" "$(patsubst qemu-%.tar.bz2,%,$@)" distclean: clean - rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi + rm -f config-host.mak config-host.h* config-host.ld $(DOCS) qemu-options.texi qemu-img-cmds.texi qemu-monitor.texi qemu-monitor-info.texi rm -f config-all-devices.mak config-all-disas.mak config.status rm -f po/*.mo tests/qemu-iotests/common.env rm -f roms/seabios/config.mak roms/vgabios/config.mak@@ -508,13 +508,16 @@ qemu-options.texi: $(SRC_PATH)/qemu-options.hx qemu-monitor.texi: $(SRC_PATH)/hmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") +qemu-monitor-info.texi: $(SRC_PATH)/hmp-commands-info.hx + $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") + qmp-commands.txt: $(SRC_PATH)/qmp-commands.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -q < $< > $@," GEN $@") qemu-img-cmds.texi: $(SRC_PATH)/qemu-img-cmds.hx $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@," GEN $@") -qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi +qemu.1: qemu-doc.texi qemu-options.texi qemu-monitor.texi qemu-monitor-info.texi $(call quiet-command, \ perl -Ww -- $(SRC_PATH)/scripts/texi2pod.pl $< qemu.pod && \ $(POD2MAN) --section=1 --center=" " --release=" " qemu.pod > $@, \@@ -551,7 +554,7 @@ pdf: qemu-doc.pdf qemu-tech.pdf qemu-doc.dvi qemu-doc.html qemu-doc.info qemu-doc.pdf: \ qemu-img.texi qemu-nbd.texi qemu-options.texi \ - qemu-monitor.texi qemu-img-cmds.texi + qemu-monitor.texi qemu-monitor-info.texi qemu-img-cmds.texi ifdef CONFIG_WIN32hmp-commands-info.hx +3
-0@@ -6,6 +6,9 @@ HXCOMM monitor info commands HXCOMM HXCOMM can be used for comments, discarded from both texi and C STEXI +@item info @var{subcommand} +@findex info +Show various information about the system state. @table @option ETEXI@@ -708,4 +711,5 @@ ETEXI
Quoted 8 lines
hmp-commands-info.hx ETEXI · line 715
Quoted 4 lines
hmp-commands-info.hx +1
-0STEXI @end table +@end table ETEXIWhere does this extra "@end table" come from?
thanks
-- PMM
This "@end table" come from hmp-commands.hx and it's necessary to close tag "@subsection Commands" from qemu-doc.texi
Re: [Qemu-devel] [PATCH 3/3] monitor: added generation of documentation for hmp-commands-info.hx
Quoted 35 lines · Pavel Butsykin
Quoted 32 lines · Peter Maydell
Quoted 22 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
It will be easier if you need to add info-commands to edit only hmp-commands-info.hx, before this had to edit
monitor.cand hmp-commands.hxhmp-commands-info.hx +3
-0@@ -6,6 +6,9 @@ HXCOMM monitor info commands HXCOMM HXCOMM can be used for comments, discarded from both texi and C STEXI +@item info @var{subcommand} +@findex info +Show various information about the system state. @table @option ETEXIhmp-commands-info.hx ETEXI · line 715
Quoted 5 lines
hmp-commands-info.hx +1
-0@@ -708,4 +711,5 @@ ETEXI STEXI @end table +@end table ETEXIWhere does this extra "@end table" come from?
Quoted 5 lines
Quoted 2 lines
thanks
-- PMMThis "@end table" come from hmp-commands.hx and it's necessary to close tag "@subsection Commands" from qemu-doc.texi
If the tag is opened in qemu-doc.texi then I think qemu-doc.texi ought to be responsible for closing it.
This may be related to a bug in the docs (as they are in master) where we close the table too early and so the 'info trace-events' and rocker related info subcommands aren't indented properly. (Also the rocker subcommands don't say "info $subcommand" the way the others do.)
thanks
-- PMM
Re: [Qemu-devel] [PATCH 3/3] monitor: added generation of documentation for hmp-commands-info.hx
Quoted 38 lines · Peter Maydell
Quoted 30 lines · Pavel Butsykin
Quoted 28 lines · Peter Maydell
Quoted 20 lines · Denis V. Lunev
From: Pavel Butsykin <[email protected]>
It will be easier if you need to add info-commands to edit only hmp-commands-info.hx, before this had to edit
monitor.cand hmp-commands.hxhmp-commands-info.hx +3
-0@@ -6,6 +6,9 @@ HXCOMM monitor info commands HXCOMM HXCOMM can be used for comments, discarded from both texi and C STEXI +@item info @var{subcommand} +@findex info +Show various information about the system state. @table @option ETEXIhmp-commands-info.hx ETEXI · line 715
Quoted 5 lines
hmp-commands-info.hx +1
-0@@ -708,4 +711,5 @@ ETEXI STEXI @end table +@end table ETEXIWhere does this extra "@end table" come from?
Quoted 4 lines
Quoted 2 lines
thanks
-- PMMThis "@end table" come from hmp-commands.hx and it's necessary to close tag "@subsection Commands" from qemu-doc.texi
If the tag is opened in qemu-doc.texi then I think qemu-doc.texi ought to be responsible for closing it.
Quoted 8 lines
This may be related to a bug in the docs (as they are in master) where we close the table too early and so the 'info trace-events' and rocker related info subcommands aren't indented properly. (Also the rocker subcommands don't say "info $subcommand" the way the others do.)
thanks
-- PMM
You're right, the "@subsection Commands" is not required to close :) I'll fix it, thanks.
Quoted 26 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.So,
monitor.calready contains a lot of generic code, as well as the target specifics code and eventuallymonitor.cvolume will only grow. This trend leads to a variety of fouling code ifdeffery(and combinations thereof), poor readability, and entanglement of architecture of the project. If someone wants to improve processing logic commands at the monitor, it isn't necessarily must differentiate amongst the implementation of some ARM orx86_64specific commands, because the project already has separation of target specific code on directories.The presented solution is not the best, but it is quite simple (PATCH doesn't add more code!) and decides the above mentioned issue. Subsequently it will not prevent the introduction of more advanced mechanism that can more effectively resolve the issue.
There is a issue with the placement of code for multiple architectures (isn't for everyone), but this code is very small. This patch is a step towards solving the issue associated with maintaining the purity of the code and structure of the project, which solves not all, but doing a little better than it is.
Quoted 9 lines
Changes from v1:
- ported to new headSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>
ping as per verbal discussion
this stuff is intrusive and become rotten very soon. Can we at least discuss it at least to come into agreement about the approach.
Den
Quoted 26 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.So,
monitor.calready contains a lot of generic code, as well as the target specifics code and eventuallymonitor.cvolume will only grow. This trend leads to a variety of fouling code ifdeffery(and combinations thereof), poor readability, and entanglement of architecture of the project. If someone wants to improve processing logic commands at the monitor, it isn't necessarily must differentiate amongst the implementation of some ARM orx86_64specific commands, because the project already has separation of target specific code on directories.The presented solution is not the best, but it is quite simple (PATCH doesn't add more code!) and decides the above mentioned issue. Subsequently it will not prevent the introduction of more advanced mechanism that can more effectively resolve the issue.
There is a issue with the placement of code for multiple architectures (isn't for everyone), but this code is very small. This patch is a step towards solving the issue associated with maintaining the purity of the code and structure of the project, which solves not all, but doing a little better than it is.
Quoted 9 lines
Changes from v1:
- ported to new headSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>
ping v2
Quoted 30 lines · Denis V. Lunev
Quoted 29 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.So,
monitor.calready contains a lot of generic code, as well as the target specifics code and eventuallymonitor.cvolume will only grow. This trend leads to a variety of fouling code ifdeffery(and combinations thereof), poor readability, and entanglement of architecture of the project. If someone wants to improve processing logic commands at the monitor, it isn't necessarily must differentiate amongst the implementation of some ARM orx86_64specific commands, because the project already has separation of target specific code on directories.The presented solution is not the best, but it is quite simple (PATCH doesn't add more code!) and decides the above mentioned issue. Subsequently it will not prevent the introduction of more advanced mechanism that can more effectively resolve the issue.
There is a issue with the placement of code for multiple architectures (isn't for everyone), but this code is very small. This patch is a step towards solving the issue associated with maintaining the purity of the code and structure of the project, which solves not all, but doing a little better than it is.
Changes from v1:
- ported to new head
Quoted 7 lines
Quoted 6 lines
Signed-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>ping v2
ping v3 as per verbal discussion on KVM forum
about timeouts :)
Quoted 10 lines · Denis V. Lunev
Quoted 9 lines · Denis V. Lunev
Quoted 7 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.
Quoted 4 lines
Quoted 1 line
ping v2
ping v3 as per verbal discussion on KVM forum
about timeouts :)
Pinging three times in a week is excessive. This is on my to-review queue, but so is a lot of other stuff.
-- PMM
Quoted 9 lines · Peter Maydell
Quoted 8 lines · Denis V. Lunev
Quoted 7 lines · Denis V. Lunev
Quoted 6 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
ping v2
ping v3 as per verbal discussion on KVM forum
about timeouts :)Pinging three times in a week is excessive. This is on my to-review queue, but so is a lot of other stuff.
-- PMM
ok. sorry for that. I'll try not to abuse you in future.
Den
P.S. Is there any public representation of your queue :) ? This may
help to avoid such misunderstandings.
Quoted 26 lines · Denis V. Lunev
The monivation of this set is simple. Recently we have proposed patch to
monitor.cwith specific x86 APIC HMP commands. The patchset was denied with the main motivation "No more arch specific code inmonitor.c" This patchset is the first step to move arch specific code frommonitor.ctargets.So,
monitor.calready contains a lot of generic code, as well as the target specifics code and eventuallymonitor.cvolume will only grow. This trend leads to a variety of fouling code ifdeffery(and combinations thereof), poor readability, and entanglement of architecture of the project. If someone wants to improve processing logic commands at the monitor, it isn't necessarily must differentiate amongst the implementation of some ARM orx86_64specific commands, because the project already has separation of target specific code on directories.The presented solution is not the best, but it is quite simple (PATCH doesn't add more code!) and decides the above mentioned issue. Subsequently it will not prevent the introduction of more advanced mechanism that can more effectively resolve the issue.
There is a issue with the placement of code for multiple architectures (isn't for everyone), but this code is very small. This patch is a step towards solving the issue associated with maintaining the purity of the code and structure of the project, which solves not all, but doing a little better than it is.
Quoted 8 lines
Changes from v1:
- ported to new headSigned-off-by Pavel Butsykin <[email protected]>Signed-off-by Denis V. Lunev <[email protected]>CC Luiz Capitulino <[email protected]>CC Paolo Bonzini <[email protected]>CC Peter Maydell <[email protected]>
Please always include diffstats in your cover letter, to make it easier for maintainers to see at a glance how much a series touches their subsystem. git-format-patch does it right:
Pavel Butsykin (3):
hmp-commands-info: move info_cmds content out of monitor.c
monitor: remove target-specific code from monitor.c
monitor: added generation of documentation for hmp-commands-info.hx
.gitignore | 1 + Makefile | 9 +- Makefile.target | 5 +- hmp-commands-info.hx | 715 ++++++++++++++++++++++ hmp-commands.hx | 120 ---- include/monitor/monitor-common.h | 43 ++ monitor.c | 1227 +------------------------------------- qemu-doc.texi | 2 + target-i386/Makefile.objs | 2 +- target-i386/monitor.c | 489 +++++++++++++++ target-ppc/Makefile.objs | 2 +- target-ppc/monitor.c | 250 ++++++++ target-sh4/Makefile.objs | 1 + target-sh4/monitor.c | 52 ++ target-sparc/Makefile.objs | 2 +- target-sparc/monitor.c | 153 +++++ target-xtensa/Makefile.objs | 1 + target-xtensa/monitor.c | 34 ++ 18 files changed, 1762 insertions(+), 1346 deletions(-) create mode 100644 hmp-commands-info.hx create mode 100644 include/monitor/monitor-common.h create mode 100644 target-i386/monitor.c create mode 100644 target-ppc/monitor.c create mode 100644 target-sh4/monitor.c create mode 100644 target-sparc/monitor.c create mode 100644 target-xtensa/monitor.c