| Message ID | 20221216221703.294683-1-allenwebb@google.com |
|---|---|
| Headers | show |
| Series | Generate modules.builtin.alias from match ids | expand |
Le 16/12/2022 à 23:16, Allen Webb a écrit : > Implement MODULE_DEVICE_TABLE for build-in modules to make it possible > to generate a builtin.alias file to complement modules.alias. > > Signed-off-by: Allen Webb <allenwebb@google.com> > --- > include/linux/module.h | 10 +++++++++- > 1 file changed, 9 insertions(+), 1 deletion(-) > > diff --git a/include/linux/module.h b/include/linux/module.h > index ec61fb53979a9..49e4019393127 100644 > --- a/include/linux/module.h > +++ b/include/linux/module.h > @@ -243,7 +243,15 @@ extern void cleanup_module(void); > extern typeof(name) __mod_##type##__##name##_device_table \ > __attribute__ ((unused, alias(__stringify(name)))) > #else /* !MODULE */ > -#define MODULE_DEVICE_TABLE(type, name) > +/* The names may not be unique for built-in modules, so include the module name > + * to guarantee uniqueness. > + */ This is network only comment style. Other parts of kenel have different style, see https://docs.kernel.org/process/coding-style.html#commenting > +#define MODULE_DEVICE_TABLE(type, name) \ > +extern void *CONCATENATE( \ 'extern' keyword is pointless of function prototypes and deprecated. Don't add new occurences. > + CONCATENATE(__mod_##type##__##name##__, \ > + __KBUILD_MODNAME), \ > + _device_table) \ > + __attribute__ ((unused, alias(__stringify(name)))) > #endif > > /* Version of form [<epoch>:]<version>[-<extra-version>].
On Sat, Dec 17, 2022 at 4:05 AM Christophe Leroy <christophe.leroy@csgroup.eu> wrote: > > > > Le 16/12/2022 à 23:16, Allen Webb a écrit : > > Implement MODULE_DEVICE_TABLE for build-in modules to make it possible > > to generate a builtin.alias file to complement modules.alias. > > > > Signed-off-by: Allen Webb <allenwebb@google.com> > > --- > > include/linux/module.h | 10 +++++++++- > > 1 file changed, 9 insertions(+), 1 deletion(-) > > > > diff --git a/include/linux/module.h b/include/linux/module.h > > index ec61fb53979a9..49e4019393127 100644 > > --- a/include/linux/module.h > > +++ b/include/linux/module.h > > @@ -243,7 +243,15 @@ extern void cleanup_module(void); > > extern typeof(name) __mod_##type##__##name##_device_table \ > > __attribute__ ((unused, alias(__stringify(name)))) > > #else /* !MODULE */ > > -#define MODULE_DEVICE_TABLE(type, name) > > +/* The names may not be unique for built-in modules, so include the module name > > + * to guarantee uniqueness. > > + */ > > This is network only comment style. I have fixed this in my local copy and will include it with the next upload. > > Other parts of kenel have different style, see > https://docs.kernel.org/process/coding-style.html#commenting > > > +#define MODULE_DEVICE_TABLE(type, name) \ > > +extern void *CONCATENATE( \ > > 'extern' keyword is pointless of function prototypes and deprecated. > Don't add new occurences. This is a weird case because these symbols are used for post compilation processing by modpost. If I drop the extern keyword, the build fails with a bunch of errors of the form: /mnt/host/source/src/third_party/kernel/upstream/drivers/hid/hid-generic.c:79:1: error: definition '__mod_hid__hid_table__kmod_hid_generic_device_table' cannot also be an alias MODULE_DEVICE_TABLE(hid, hid_table); ^ /mnt/host/source/src/third_party/kernel/upstream/include/linux/module.h:255:26: note: expanded from macro 'MODULE_DEVICE_TABLE' __attribute__ ((unused, alias(__stringify(name)))) > > > > + CONCATENATE(__mod_##type##__##name##__, \ > > + __KBUILD_MODNAME), \ > > + _device_table) \ > > + __attribute__ ((unused, alias(__stringify(name)))) > > #endif > > > > /* Version of form [<epoch>:]<version>[-<extra-version>].
On Mon, Dec 19, 2022 at 01:18:47PM -0600, Allen Webb wrote: > A one character difference in the name supplied to MODULE_DEVICE_TABLE > breaks a future patch set, so fix the typo. > > Reported-by: kernel test robot <lkp@intel.com> > Signed-off-by: Allen Webb <allenwebb@google.com> > --- > drivers/soc/imx/imx8mp-blk-ctrl.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/soc/imx/imx8mp-blk-ctrl.c b/drivers/soc/imx/imx8mp-blk-ctrl.c > index 0e3b6ba22f943..344a0a71df14a 100644 > --- a/drivers/soc/imx/imx8mp-blk-ctrl.c > +++ b/drivers/soc/imx/imx8mp-blk-ctrl.c > @@ -743,7 +743,7 @@ static const struct of_device_id imx8mp_blk_ctrl_of_match[] = { > /* Sentinel */ > } > }; > -MODULE_DEVICE_TABLE(of, imx8m_blk_ctrl_of_match); > +MODULE_DEVICE_TABLE(of, imx8mp_blk_ctrl_of_match); What commit id does this fix? Shouldn't this be also cc: stable to resolve this issue for older kernels as obviousl the module device table for auto-loading is not correct? thanks, greg k-h
On Mon, Dec 19, 2022 at 01:18:46PM -0600, Allen Webb wrote:
> Generate modules.builtin.alias from match ids
This is looking much better, thanks! Please expand with proper
documentation on the use case / value of this on the file:
Documentation/kbuild/kbuild.rst
Luis