Hello Guang,
Are you okay with this version?
Regards,
Andreas
v5 -> v6:
* Fix bisectability by deferring use of hw_{local,global}_breakpoint_enabled().
* Reword the commit messages for clarity.
* Squash more constant usage into first patch.
* Reorder DATA_RW and DATA_WR to match original switch cases.
* Untangle introduction of breakpoint helper functions from goto/if
refactorings.
* Make hw_breakpoint_enabled() return bool.
* Keep IO_RW in place for patch readability.
* Keep reg variable name for patch readability.
* Move {bp,wp}_match inside loop to clarify scope and to avoid double false
init.
* Make check_hw_breakpoints() return bool, change force_dr6_update arg to bool.
changes v4->v5:
- fix some not well formated changes.
- split functional and non-functional changes for cherry-picking
suggested by Andreas Färber <afaer...@suse.de>
changes v3->v4:
- fix wrong logic of hw_{global,local}_breakpoint_enabled usage
suggested by Peter Maydell <peter.mayd...@linaro.org>
changes v2->v3:
- split hw_breakpoint_enabled into hw_{global,local}_breakpoint_enabled
changes v1->v2:
- add _TYPE_ to the name of dr7 bit field
- fix some coding styles
suggested by Peter Maydell <peter.mayd...@linaro.org>
Cc: liguang <lig.f...@cn.fujitsu.com>
Cc: Eduardo Habkost <ehabk...@redhat.com>
Cc: Igor Mammedov <imamm...@redhat.com>
Cc: Jan Kiszka <jan.kis...@siemens.com>
Cc: Peter Maydell <peter.mayd...@linaro.org>
liguang (4):
target-i386: Define DR7 bit field constants
target-i386: Introduce hw_{local,global}_breakpoint_enabled()
target-i386: Avoid goto in hw_breakpoint_insert()
target-i386: Use switch in check_hw_breakpoints()
target-i386/cpu.h | 23 ++++++++++--
target-i386/helper.c | 87 +++++++++++++++++++++++++++++----------------
target-i386/machine.c | 5 +--
target-i386/misc_helper.c | 6 ++--
target-i386/seg_helper.c | 9 ++---
5 Dateien geändert, 88 Zeilen hinzugefügt(+), 42 Zeilen entfernt(-)
--
1.7.10.4
