Difference between select and enable in Kconfig
Hi,
while porting some of my drivers to 2.6 which use the firmware loader for example I came to the question whether to use select or enable to achieve the desired result. Looking at the documention don't gives me the answer and from zconf.l I feel that both options are the same. Can anyone please explain me the differences if there are any?
Regards
Marcel
On Sun, 21 Dec 2003 03:46:54 +0100 Marcel Holtmann <[email protected]> wrote:
| Hi, | | while porting some of my drivers to 2.6 which use the firmware loader | for example I came to the question whether to use select or enable to | achieve the desired result. Looking at the documention don't gives me | the answer and from zconf.l I feel that both options are the same. Can | anyone please explain me the differences if there are any?
I agree, they look like synonyms.
There's nothing about "enable" in Documentation/kbuild/kconfig-language.txt and there's evidence in zconf.l (as you mention) and in menu.c that they are the same:
case P_SELECT: sym2 = prop_get_symbol(prop); if ((sym->type != S_BOOLEAN && sym->type != S_TRISTATE) || (sym2->type != S_BOOLEAN && sym2->type != S_TRISTATE)) fprintf(stderr, "%s:%d:warning: enable is only allowed with boolean and tristate symbols\n", ~~~~~~
--
~Randy
Hi Randy,
Quoted 11 lines
| while porting some of my drivers to 2.6 which use the firmware loader | for example I came to the question whether to use select or enable to | achieve the desired result. Looking at the documention don't gives me | the answer and from zconf.l I feel that both options are the same. Can | anyone please explain me the differences if there are any?I agree, they look like synonyms.
There's nothing about "enable" inDocumentation/kbuild/kconfig-language.txtand there's evidence in zconf.l (as you mention) and inmenu.cthat they are the same:
Quoted 6 lines
case P_SELECT: sym2 = prop_get_symbol(prop); if ((sym->type != S_BOOLEAN && sym->type != S_TRISTATE) || (sym2->type != S_BOOLEAN && sym2->type != S_TRISTATE)) fprintf(stderr, "%s:%d:warning: enable is only allowed with boolean and tristate symbols\n", ~~~~~~
so both options achieve the same result. Why do we have two different options for the same stuff? Should we not remove one?
Regards
Marcel
Hi,
Quoted 9 lines · Marcel Holtmann
Quoted 6 lines
case P_SELECT: sym2 = prop_get_symbol(prop); if ((sym->type != S_BOOLEAN && sym->type != S_TRISTATE) || (sym2->type != S_BOOLEAN && sym2->type != S_TRISTATE)) fprintf(stderr, "%s:%d:warning: enable is only allowed with boolean and tristate symbols\n", ~~~~~~so both options achieve the same result. Why do we have two different options for the same stuff? Should we not remove one?
It was called first 'enable' and later renamed into 'select', which is now the official version, so 'enable' could be indeed removed.
bye, Roman
Hi Roman,
Quoted 5 lines
Quoted 2 lines
so both options achieve the same result. Why do we have two different options for the same stuff? Should we not remove one?
It was called first 'enable' and later renamed into 'select', which is now the official version, so 'enable' could be indeed removed.
thanks for the clarification and I have choosen the wrong one. After a quick grep through the kernel sources I found that
drivers/net/wireless/Kconfig
is the only file that uses the enable option and of course this was the example I have looked at :(
Regards
Marcel