| Message ID | 1515567374-12722-3-git-send-email-yamada.masahiro@socionext.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Jan 10, 2018 at 7:56 AM, Masahiro Yamada <yamada.masahiro@socionext.com> wrote: > (all)defconfig does not need to call conf_set_all_new_symbols() because > conf_write() calculates all symbols based on their default unless they > have been user-defined. > > conf_set_all_new_symbols(def_default) is no-op except for "choice". > It calls sym_cal_value() for "choice", but the SYMBOL_VALID is cleared > by set_all_choice_values() later on. So, conf_write() will re-calculate > it, which is just a wasteful computation. The only difference I see > is SYMBOL_DEF_USER flag for "choice", but it gives no difference to the > resulted .config file. > > I confirmed this change still produced the same .config for all ARCH. > > I also moved savedefconfig to collect the no-op cases at the end of > the switch statement. > > Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> > --- > > scripts/kconfig/conf.c | 11 +++-------- > 1 file changed, 3 insertions(+), 8 deletions(-) > > diff --git a/scripts/kconfig/conf.c b/scripts/kconfig/conf.c > index 52cbe5d..8364811 100644 > --- a/scripts/kconfig/conf.c > +++ b/scripts/kconfig/conf.c > @@ -655,18 +655,10 @@ int main(int ac, char **av) > case allmodconfig: > conf_set_all_new_symbols(def_mod); > break; > - case alldefconfig: > - conf_set_all_new_symbols(def_default); > - break; > case randconfig: > /* Really nothing to do in this loop */ > while (conf_set_all_new_symbols(def_random)) ; > break; > - case defconfig: > - conf_set_all_new_symbols(def_default); > - break; > - case savedefconfig: > - break; > case oldaskconfig: > rootEntry = &rootmenu; > conf(&rootmenu); > @@ -681,7 +673,10 @@ int main(int ac, char **av) > check_conf(&rootmenu); > } while (conf_cnt && input_mode != listnewconfig); > break; > + case alldefconfig: > + case defconfig: > case olddefconfig: > + case savedefconfig: > default: > break; > } > -- > 2.7.4 > Another side effect of conf_set_all_new_symbols(def_default) is that SYMBOL_NEED_SET_CHOICE_VALUES is set on all choice values. Might want to check fbe98bb9ed3d ("kconfig: Fix defconfig when one choice menu selects options that another choice menu depends on") to make sure nothing breaks related to that. Cheers, Ulf -- To unsubscribe from this list: send the line "unsubscribe linux-kbuild" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Ulf, 2018-01-12 23:28 GMT+09:00 Ulf Magnusson <ulfalizer@gmail.com>: > On Wed, Jan 10, 2018 at 7:56 AM, Masahiro Yamada > <yamada.masahiro@socionext.com> wrote: >> (all)defconfig does not need to call conf_set_all_new_symbols() because >> conf_write() calculates all symbols based on their default unless they >> have been user-defined. >> >> conf_set_all_new_symbols(def_default) is no-op except for "choice". >> It calls sym_cal_value() for "choice", but the SYMBOL_VALID is cleared >> by set_all_choice_values() later on. So, conf_write() will re-calculate >> it, which is just a wasteful computation. The only difference I see >> is SYMBOL_DEF_USER flag for "choice", but it gives no difference to the >> resulted .config file. >> >> I confirmed this change still produced the same .config for all ARCH. >> >> I also moved savedefconfig to collect the no-op cases at the end of >> the switch statement. >> >> Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> >> --- >> >> scripts/kconfig/conf.c | 11 +++-------- >> 1 file changed, 3 insertions(+), 8 deletions(-) >> >> diff --git a/scripts/kconfig/conf.c b/scripts/kconfig/conf.c >> index 52cbe5d..8364811 100644 >> --- a/scripts/kconfig/conf.c >> +++ b/scripts/kconfig/conf.c >> @@ -655,18 +655,10 @@ int main(int ac, char **av) >> case allmodconfig: >> conf_set_all_new_symbols(def_mod); >> break; >> - case alldefconfig: >> - conf_set_all_new_symbols(def_default); >> - break; >> case randconfig: >> /* Really nothing to do in this loop */ >> while (conf_set_all_new_symbols(def_random)) ; >> break; >> - case defconfig: >> - conf_set_all_new_symbols(def_default); >> - break; >> - case savedefconfig: >> - break; >> case oldaskconfig: >> rootEntry = &rootmenu; >> conf(&rootmenu); >> @@ -681,7 +673,10 @@ int main(int ac, char **av) >> check_conf(&rootmenu); >> } while (conf_cnt && input_mode != listnewconfig); >> break; >> + case alldefconfig: >> + case defconfig: >> case olddefconfig: >> + case savedefconfig: >> default: >> break; >> } >> -- >> 2.7.4 >> > > Another side effect of conf_set_all_new_symbols(def_default) is that > SYMBOL_NEED_SET_CHOICE_VALUES is set on all choice values. Might want > to check fbe98bb9ed3d ("kconfig: Fix defconfig when one choice menu > selects options that another choice menu depends on") to make sure > nothing breaks related to that. > > Cheers, > Ulf > -- Thanks for your comments! From a quick test, this patch seems OK, but I need to take a closer look at this.
diff --git a/scripts/kconfig/conf.c b/scripts/kconfig/conf.c index 52cbe5d..8364811 100644 --- a/scripts/kconfig/conf.c +++ b/scripts/kconfig/conf.c @@ -655,18 +655,10 @@ int main(int ac, char **av) case allmodconfig: conf_set_all_new_symbols(def_mod); break; - case alldefconfig: - conf_set_all_new_symbols(def_default); - break; case randconfig: /* Really nothing to do in this loop */ while (conf_set_all_new_symbols(def_random)) ; break; - case defconfig: - conf_set_all_new_symbols(def_default); - break; - case savedefconfig: - break; case oldaskconfig: rootEntry = &rootmenu; conf(&rootmenu); @@ -681,7 +673,10 @@ int main(int ac, char **av) check_conf(&rootmenu); } while (conf_cnt && input_mode != listnewconfig); break; + case alldefconfig: + case defconfig: case olddefconfig: + case savedefconfig: default: break; }
(all)defconfig does not need to call conf_set_all_new_symbols() because conf_write() calculates all symbols based on their default unless they have been user-defined. conf_set_all_new_symbols(def_default) is no-op except for "choice". It calls sym_cal_value() for "choice", but the SYMBOL_VALID is cleared by set_all_choice_values() later on. So, conf_write() will re-calculate it, which is just a wasteful computation. The only difference I see is SYMBOL_DEF_USER flag for "choice", but it gives no difference to the resulted .config file. I confirmed this change still produced the same .config for all ARCH. I also moved savedefconfig to collect the no-op cases at the end of the switch statement. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> --- scripts/kconfig/conf.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-)