2019-03-15 10:11:20

by Alexey Gladkov

[permalink] [raw]
Subject: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

Problem:

When a kernel module is compiled as a separate module, some important
information about the kernel module is available via .modinfo section of
the module. In contrast, when the kernel module is compiled into the
kernel, that information is not available.

Information about built-in modules is necessary in the following cases:

1. When it is necessary to find out what additional parameters can be
passed to the kernel at boot time.

2. When you need to know which module names and their aliases are in
the kernel. This is very useful for creating an initrd image.

Proposal:

The proposed patch does not remove .modinfo section with module
information from the vmlinux at the build time and saves it into a
separate file after kernel linking. So, the kernel does not increase in
size and no additional information remains in it. Information is stored
in the same format as in the separate modules (null-terminated string
array). Because the .modinfo section is already exported with a separate
modules, we are not creating a new API.

It can be easily read in the userspace:

$ tr '\0' '\n' < kernel.builtin.modinfo
ext4.softdep=pre: crc32c
ext4.license=GPL
ext4.description=Fourth Extended Filesystem
ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
ext4.alias=fs-ext4
ext4.alias=ext3
ext4.alias=fs-ext3
ext4.alias=ext2
ext4.alias=fs-ext2
md_mod.alias=block-major-9-*
md_mod.alias=md
md_mod.description=MD RAID framework
md_mod.license=GPL
md_mod.parmtype=create_on_open:bool
md_mod.parmtype=start_dirty_degraded:int
...

Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
Signed-off-by: Alexey Gladkov <[email protected]>
---
Makefile | 1 +
include/linux/moduleparam.h | 12 +++++-------
scripts/link-vmlinux.sh | 8 ++++++++
3 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/Makefile b/Makefile
index d5713e7b1e50..971102194c92 100644
--- a/Makefile
+++ b/Makefile
@@ -1288,6 +1288,7 @@ _modinst_:
fi
@cp -f $(objtree)/modules.order $(MODLIB)/
@cp -f $(objtree)/modules.builtin $(MODLIB)/
+ @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/
$(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst

# This depmod is only for convenience to give the initial
diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
index ba36506db4fb..5ba250d9172a 100644
--- a/include/linux/moduleparam.h
+++ b/include/linux/moduleparam.h
@@ -10,23 +10,21 @@
module name. */
#ifdef MODULE
#define MODULE_PARAM_PREFIX /* empty */
+#define __MODULE_INFO_PREFIX /* empty */
#else
#define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
+/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
+#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
#endif

/* Chosen so that structs with an unsigned long line up. */
#define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))

-#ifdef MODULE
#define __MODULE_INFO(tag, name, info) \
static const char __UNIQUE_ID(name)[] \
__used __attribute__((section(".modinfo"), unused, aligned(1))) \
- = __stringify(tag) "=" info
-#else /* !MODULE */
-/* This struct is here for syntactic coherency, it is not used */
-#define __MODULE_INFO(tag, name, info) \
- struct __UNIQUE_ID(name) {}
-#endif
+ = __MODULE_INFO_PREFIX __stringify(tag) "=" info
+
#define __MODULE_PARM_TYPE(name, _type) \
__MODULE_INFO(parmtype, name##type, #name ":" _type)

diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
index c8cf45362bd6..399d7e4d11ec 100755
--- a/scripts/link-vmlinux.sh
+++ b/scripts/link-vmlinux.sh
@@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then

# step 1
vmlinux_link "" .tmp_vmlinux1
+ "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o

# step 2
vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
+ "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o

# step 3
@@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
kallsyms_vmlinux=.tmp_vmlinux3

vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
+ "${OBJCOPY}" -R .modinfo .tmp_vmlinux3

kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
fi
@@ -281,6 +284,11 @@ fi
info LD vmlinux
vmlinux_link "${kallsymso}" vmlinux

+info MODINFO kernel.builtin.modinfo
+"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
+chmod 444 kernel.builtin.modinfo
+"${OBJCOPY}" -R .modinfo vmlinux
+
if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
info SORTEX vmlinux
sortextable vmlinux
--
2.19.2



2019-03-20 08:56:41

by Masahiro Yamada

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
>
> Problem:
>
> When a kernel module is compiled as a separate module, some important
> information about the kernel module is available via .modinfo section of
> the module. In contrast, when the kernel module is compiled into the
> kernel, that information is not available.
>
> Information about built-in modules is necessary in the following cases:
>
> 1. When it is necessary to find out what additional parameters can be
> passed to the kernel at boot time.
>
> 2. When you need to know which module names and their aliases are in
> the kernel. This is very useful for creating an initrd image.
>
> Proposal:
>
> The proposed patch does not remove .modinfo section with module
> information from the vmlinux at the build time and saves it into a
> separate file after kernel linking. So, the kernel does not increase in
> size and no additional information remains in it. Information is stored
> in the same format as in the separate modules (null-terminated string
> array). Because the .modinfo section is already exported with a separate
> modules, we are not creating a new API.
>
> It can be easily read in the userspace:
>
> $ tr '\0' '\n' < kernel.builtin.modinfo
> ext4.softdep=pre: crc32c
> ext4.license=GPL
> ext4.description=Fourth Extended Filesystem
> ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
> ext4.alias=fs-ext4
> ext4.alias=ext3
> ext4.alias=fs-ext3
> ext4.alias=ext2
> ext4.alias=fs-ext2
> md_mod.alias=block-major-9-*
> md_mod.alias=md
> md_mod.description=MD RAID framework
> md_mod.license=GPL
> md_mod.parmtype=create_on_open:bool
> md_mod.parmtype=start_dirty_degraded:int
> ...
>
> Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> Signed-off-by: Alexey Gladkov <[email protected]>
> ---


I just started to take a look.

I was able to compile it for x86,
but got tons of warnings for powerpc.

Anyway, you do not need to send v2 soon.
Please give me some more time for the through review.



The following is what I got by
"make -j8 ARCH=powerpc CROSS_COMPILE=powerpc-linux- defconfig all"


...
CC [M] drivers/scsi/st.o
AR drivers/scsi/built-in.a
AR drivers/built-in.a
GEN .version
CHK include/generated/compile.h
UPD include/generated/compile.h
CC init/version.o
AR init/built-in.a
LD vmlinux.o
MODPOST vmlinux.o
powerpc-linux-ld: warning: orphan section `.gnu.hash' from `linker
stubs' being placed in section `.gnu.hash'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`arch/powerpc/platforms/pseries/pseries_energy.o' being placed in
section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`arch/powerpc/platforms/pseries/cmm.o' being placed in section
`.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`arch/powerpc/platforms/pasemi/gpio_mdio.o' being placed in section
`.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/workqueue.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/printk/printk.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/irq/spurious.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/rcu/update.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/rcu/srcutree.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/rcu/tree.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/module.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`kernel/configs.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/binfmt_script.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/binfmt_elf.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/compat_binfmt_elf.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/mbcache.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs_common/nfsacl.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs_common/grace.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/ext4/super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/ext2/super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/jbd2/journal.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/fat/inode.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/fat/namei_msdos.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/isofs/inode.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/dir.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/inode.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/nfs2super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/nfs3super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nfs/nfs4super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/exportfs/expfs.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/lockd/svc.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nls/nls_base.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nls/nls_cp437.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nls/nls_ascii.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nls/nls_iso8859-1.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/nls/nls_utf8.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/xfs/xfs_super.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/debugfs/inode.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`fs/tracefs/inode.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/api.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/crypto_wq.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/algapi.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/aead.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/blkcipher.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/skcipher.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/ahash.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/shash.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/akcipher.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/kpp.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/acompress.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/scompress.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/algboss.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/testmgr.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/hmac.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/crypto_null.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/sha256_generic.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/aes_generic.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/crc32c_generic.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`crypto/rng.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`block/bsg.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`block/mq-deadline.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`block/kyber-iosched.o' being placed in section `.modinfo'.
powerpc-linux-ld: warning: orphan section `.modinfo' from
`lib/btree.o' being placed in section `.modinfo'.

...





> Makefile | 1 +
> include/linux/moduleparam.h | 12 +++++-------
> scripts/link-vmlinux.sh | 8 ++++++++
> 3 files changed, 14 insertions(+), 7 deletions(-)
>
> diff --git a/Makefile b/Makefile
> index d5713e7b1e50..971102194c92 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -1288,6 +1288,7 @@ _modinst_:
> fi
> @cp -f $(objtree)/modules.order $(MODLIB)/
> @cp -f $(objtree)/modules.builtin $(MODLIB)/
> + @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/
> $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
>
> # This depmod is only for convenience to give the initial
> diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
> index ba36506db4fb..5ba250d9172a 100644
> --- a/include/linux/moduleparam.h
> +++ b/include/linux/moduleparam.h
> @@ -10,23 +10,21 @@
> module name. */
> #ifdef MODULE
> #define MODULE_PARAM_PREFIX /* empty */
> +#define __MODULE_INFO_PREFIX /* empty */
> #else
> #define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
> +/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
> +#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
> #endif
>
> /* Chosen so that structs with an unsigned long line up. */
> #define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
>
> -#ifdef MODULE
> #define __MODULE_INFO(tag, name, info) \
> static const char __UNIQUE_ID(name)[] \
> __used __attribute__((section(".modinfo"), unused, aligned(1))) \
> - = __stringify(tag) "=" info
> -#else /* !MODULE */
> -/* This struct is here for syntactic coherency, it is not used */
> -#define __MODULE_INFO(tag, name, info) \
> - struct __UNIQUE_ID(name) {}
> -#endif
> + = __MODULE_INFO_PREFIX __stringify(tag) "=" info
> +
> #define __MODULE_PARM_TYPE(name, _type) \
> __MODULE_INFO(parmtype, name##type, #name ":" _type)
>
> diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> index c8cf45362bd6..399d7e4d11ec 100755
> --- a/scripts/link-vmlinux.sh
> +++ b/scripts/link-vmlinux.sh
> @@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
>
> # step 1
> vmlinux_link "" .tmp_vmlinux1
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
> kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o
>
> # step 2
> vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
> kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o
>
> # step 3
> @@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> kallsyms_vmlinux=.tmp_vmlinux3
>
> vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux3
>
> kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
> fi
> @@ -281,6 +284,11 @@ fi
> info LD vmlinux
> vmlinux_link "${kallsymso}" vmlinux
>
> +info MODINFO kernel.builtin.modinfo
> +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
> +chmod 444 kernel.builtin.modinfo
> +"${OBJCOPY}" -R .modinfo vmlinux
> +
> if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
> info SORTEX vmlinux
> sortextable vmlinux
> --
> 2.19.2
>


--
Best Regards
Masahiro Yamada

2019-03-22 05:35:55

by Masahiro Yamada

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

Hi.

(added some people to CC)


On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
>
> Problem:
>
> When a kernel module is compiled as a separate module, some important
> information about the kernel module is available via .modinfo section of
> the module. In contrast, when the kernel module is compiled into the
> kernel, that information is not available.


I might be missing something, but
vmlinux provides info of builtin modules
in /sys/module/.

(Looks like currently only module_param and MODULE_VERSION)

This patch is not exactly the same, but I see a kind of overwrap.
I'd like to be sure if we want this new scheme.


> Information about built-in modules is necessary in the following cases:
>
> 1. When it is necessary to find out what additional parameters can be
> passed to the kernel at boot time.


Actually, /sys/module/<module>/parameters/
exposes this information.

Doesn't it work for your purpose?



> 2. When you need to know which module names and their aliases are in
> the kernel. This is very useful for creating an initrd image.
>
> Proposal:
>
> The proposed patch does not remove .modinfo section with module
> information from the vmlinux at the build time and saves it into a
> separate file after kernel linking. So, the kernel does not increase in
> size and no additional information remains in it. Information is stored
> in the same format as in the separate modules (null-terminated string
> array). Because the .modinfo section is already exported with a separate
> modules, we are not creating a new API.
>
> It can be easily read in the userspace:
>
> $ tr '\0' '\n' < kernel.builtin.modinfo
> ext4.softdep=pre: crc32c
> ext4.license=GPL
> ext4.description=Fourth Extended Filesystem
> ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
> ext4.alias=fs-ext4
> ext4.alias=ext3
> ext4.alias=fs-ext3
> ext4.alias=ext2
> ext4.alias=fs-ext2
> md_mod.alias=block-major-9-*
> md_mod.alias=md
> md_mod.description=MD RAID framework
> md_mod.license=GPL
> md_mod.parmtype=create_on_open:bool
> md_mod.parmtype=start_dirty_degraded:int
> ...
>
> Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> Signed-off-by: Alexey Gladkov <[email protected]>
> ---
> Makefile | 1 +
> include/linux/moduleparam.h | 12 +++++-------
> scripts/link-vmlinux.sh | 8 ++++++++
> 3 files changed, 14 insertions(+), 7 deletions(-)
>
> diff --git a/Makefile b/Makefile
> index d5713e7b1e50..971102194c92 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -1288,6 +1288,7 @@ _modinst_:
> fi
> @cp -f $(objtree)/modules.order $(MODLIB)/
> @cp -f $(objtree)/modules.builtin $(MODLIB)/
> + @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/
> $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
>
> # This depmod is only for convenience to give the initial
> diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
> index ba36506db4fb..5ba250d9172a 100644
> --- a/include/linux/moduleparam.h
> +++ b/include/linux/moduleparam.h
> @@ -10,23 +10,21 @@
> module name. */
> #ifdef MODULE
> #define MODULE_PARAM_PREFIX /* empty */
> +#define __MODULE_INFO_PREFIX /* empty */
> #else
> #define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
> +/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
> +#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
> #endif
>
> /* Chosen so that structs with an unsigned long line up. */
> #define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
>
> -#ifdef MODULE
> #define __MODULE_INFO(tag, name, info) \
> static const char __UNIQUE_ID(name)[] \
> __used __attribute__((section(".modinfo"), unused, aligned(1))) \
> - = __stringify(tag) "=" info
> -#else /* !MODULE */
> -/* This struct is here for syntactic coherency, it is not used */
> -#define __MODULE_INFO(tag, name, info) \
> - struct __UNIQUE_ID(name) {}
> -#endif
> + = __MODULE_INFO_PREFIX __stringify(tag) "=" info
> +
> #define __MODULE_PARM_TYPE(name, _type) \
> __MODULE_INFO(parmtype, name##type, #name ":" _type)
>
> diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> index c8cf45362bd6..399d7e4d11ec 100755
> --- a/scripts/link-vmlinux.sh
> +++ b/scripts/link-vmlinux.sh
> @@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
>
> # step 1
> vmlinux_link "" .tmp_vmlinux1
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
> kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o
>
> # step 2
> vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
> kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o
>
> # step 3
> @@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> kallsyms_vmlinux=.tmp_vmlinux3
>
> vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
> + "${OBJCOPY}" -R .modinfo .tmp_vmlinux3
>
> kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
> fi
> @@ -281,6 +284,11 @@ fi
> info LD vmlinux
> vmlinux_link "${kallsymso}" vmlinux
>
> +info MODINFO kernel.builtin.modinfo
> +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
> +chmod 444 kernel.builtin.modinfo
> +"${OBJCOPY}" -R .modinfo vmlinux
> +
> if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
> info SORTEX vmlinux
> sortextable vmlinux
> --
> 2.19.2
>


--
Best Regards
Masahiro Yamada

2019-03-26 17:25:57

by Alexey Gladkov

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> Hi.
>
> (added some people to CC)
>
>
> On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> >
> > Problem:
> >
> > When a kernel module is compiled as a separate module, some important
> > information about the kernel module is available via .modinfo section of
> > the module. In contrast, when the kernel module is compiled into the
> > kernel, that information is not available.
>
>
> I might be missing something, but
> vmlinux provides info of builtin modules
> in /sys/module/.

No. There are definitely not all modules. I have a builtin sha256_generic,
but I can't find him in the /sys/module.

> (Looks like currently only module_param and MODULE_VERSION)
>
> This patch is not exactly the same, but I see a kind of overwrap.
> I'd like to be sure if we want this new scheme.

The /sys/module is only for running kernel. One of my use cases is
to create an initrd for a new kernel.

>
> > Information about built-in modules is necessary in the following cases:
> >
> > 1. When it is necessary to find out what additional parameters can be
> > passed to the kernel at boot time.
>
>
> Actually, /sys/module/<module>/parameters/
> exposes this information.
>
> Doesn't it work for your purpose?

No, since creating an initrd needs to know all the modalias before
I get the sysfs for new kernel. Also there are no modalias at all.

> > 2. When you need to know which module names and their aliases are in
> > the kernel. This is very useful for creating an initrd image.
> >
> > Proposal:
> >
> > The proposed patch does not remove .modinfo section with module
> > information from the vmlinux at the build time and saves it into a
> > separate file after kernel linking. So, the kernel does not increase in
> > size and no additional information remains in it. Information is stored
> > in the same format as in the separate modules (null-terminated string
> > array). Because the .modinfo section is already exported with a separate
> > modules, we are not creating a new API.
> >
> > It can be easily read in the userspace:
> >
> > $ tr '\0' '\n' < kernel.builtin.modinfo
> > ext4.softdep=pre: crc32c
> > ext4.license=GPL
> > ext4.description=Fourth Extended Filesystem
> > ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
> > ext4.alias=fs-ext4
> > ext4.alias=ext3
> > ext4.alias=fs-ext3
> > ext4.alias=ext2
> > ext4.alias=fs-ext2
> > md_mod.alias=block-major-9-*
> > md_mod.alias=md
> > md_mod.description=MD RAID framework
> > md_mod.license=GPL
> > md_mod.parmtype=create_on_open:bool
> > md_mod.parmtype=start_dirty_degraded:int
> > ...
> >
> > Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > Signed-off-by: Alexey Gladkov <[email protected]>
> > ---
> > Makefile | 1 +
> > include/linux/moduleparam.h | 12 +++++-------
> > scripts/link-vmlinux.sh | 8 ++++++++
> > 3 files changed, 14 insertions(+), 7 deletions(-)
> >
> > diff --git a/Makefile b/Makefile
> > index d5713e7b1e50..971102194c92 100644
> > --- a/Makefile
> > +++ b/Makefile
> > @@ -1288,6 +1288,7 @@ _modinst_:
> > fi
> > @cp -f $(objtree)/modules.order $(MODLIB)/
> > @cp -f $(objtree)/modules.builtin $(MODLIB)/
> > + @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/
> > $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
> >
> > # This depmod is only for convenience to give the initial
> > diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
> > index ba36506db4fb..5ba250d9172a 100644
> > --- a/include/linux/moduleparam.h
> > +++ b/include/linux/moduleparam.h
> > @@ -10,23 +10,21 @@
> > module name. */
> > #ifdef MODULE
> > #define MODULE_PARAM_PREFIX /* empty */
> > +#define __MODULE_INFO_PREFIX /* empty */
> > #else
> > #define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
> > +/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
> > +#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
> > #endif
> >
> > /* Chosen so that structs with an unsigned long line up. */
> > #define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
> >
> > -#ifdef MODULE
> > #define __MODULE_INFO(tag, name, info) \
> > static const char __UNIQUE_ID(name)[] \
> > __used __attribute__((section(".modinfo"), unused, aligned(1))) \
> > - = __stringify(tag) "=" info
> > -#else /* !MODULE */
> > -/* This struct is here for syntactic coherency, it is not used */
> > -#define __MODULE_INFO(tag, name, info) \
> > - struct __UNIQUE_ID(name) {}
> > -#endif
> > + = __MODULE_INFO_PREFIX __stringify(tag) "=" info
> > +
> > #define __MODULE_PARM_TYPE(name, _type) \
> > __MODULE_INFO(parmtype, name##type, #name ":" _type)
> >
> > diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> > index c8cf45362bd6..399d7e4d11ec 100755
> > --- a/scripts/link-vmlinux.sh
> > +++ b/scripts/link-vmlinux.sh
> > @@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> >
> > # step 1
> > vmlinux_link "" .tmp_vmlinux1
> > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
> > kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o
> >
> > # step 2
> > vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
> > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
> > kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o
> >
> > # step 3
> > @@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> > kallsyms_vmlinux=.tmp_vmlinux3
> >
> > vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
> > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux3
> >
> > kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
> > fi
> > @@ -281,6 +284,11 @@ fi
> > info LD vmlinux
> > vmlinux_link "${kallsymso}" vmlinux
> >
> > +info MODINFO kernel.builtin.modinfo
> > +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
> > +chmod 444 kernel.builtin.modinfo
> > +"${OBJCOPY}" -R .modinfo vmlinux
> > +
> > if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
> > info SORTEX vmlinux
> > sortextable vmlinux
> > --
> > 2.19.2
> >
>
>
> --
> Best Regards
> Masahiro Yamada
>

--
Rgrds, legion


2019-03-27 15:41:43

by Jessica Yu

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

+++ Alexey Gladkov [26/03/19 18:24 +0100]:
>On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
>> Hi.
>>
>> (added some people to CC)

(Thanks Masahiro for the CC!)

>>
>> On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
>> >
>> > Problem:
>> >
>> > When a kernel module is compiled as a separate module, some important
>> > information about the kernel module is available via .modinfo section of
>> > the module. In contrast, when the kernel module is compiled into the
>> > kernel, that information is not available.
>>
>>
>> I might be missing something, but
>> vmlinux provides info of builtin modules
>> in /sys/module/.
>
>No. There are definitely not all modules. I have a builtin sha256_generic,
>but I can't find him in the /sys/module.

Yeah, you'll only find builtin modules under /sys/module/ if it has any module
parameters, otherwise you won't find it there. As Masahiro already mentioned,
if a builtin module has any parameters, they would be accessible under /sys/module/.

>> (Looks like currently only module_param and MODULE_VERSION)
>>
>> This patch is not exactly the same, but I see a kind of overwrap.
>> I'd like to be sure if we want this new scheme.
>
>The /sys/module is only for running kernel. One of my use cases is
>to create an initrd for a new kernel.
>
>>
>> > Information about built-in modules is necessary in the following cases:
>> >
>> > 1. When it is necessary to find out what additional parameters can be
>> > passed to the kernel at boot time.
>>
>>
>> Actually, /sys/module/<module>/parameters/
>> exposes this information.
>>
>> Doesn't it work for your purpose?
>
>No, since creating an initrd needs to know all the modalias before
>I get the sysfs for new kernel. Also there are no modalias at all.
>
>> > 2. When you need to know which module names and their aliases are in
>> > the kernel. This is very useful for creating an initrd image.
>> >

Hm, I do see one possible additional use-case for preserving module alias
information for built-in modules - modprobe will currently error (I think,
correct me if I'm wrong) if we try invoking modprobe with an alias of a
built-in module, simply because this information is not in modules.builtin or
modules.alias.

Since kbuild already outputs modules.builtin, I would suggest outputting
something like a modules.builtin.alias file (and I guess maybe a modules.builtin.param
file too if that's deemed useful), in a format that is consumable by kmod/modprobe,
so that modprobing an alias of a built-in module doesn't produce an error. I
think this should be easy to do if we keep and parse the resulting .modinfo for
builtin modules. This is just an idea, opinions welcome. I've added Lucas to CC
in case he has any thoughts.


Thanks,

Jessica


2019-03-27 16:05:58

by Alexey Gladkov

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Wed, Mar 27, 2019 at 04:40:25PM +0100, Jessica Yu wrote:
> +++ Alexey Gladkov [26/03/19 18:24 +0100]:
> >On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> >> Hi.
> >>
> >> (added some people to CC)
>
> (Thanks Masahiro for the CC!)
>
> >>
> >> On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> >> >
> >> > Problem:
> >> >
> >> > When a kernel module is compiled as a separate module, some important
> >> > information about the kernel module is available via .modinfo section of
> >> > the module. In contrast, when the kernel module is compiled into the
> >> > kernel, that information is not available.
> >>
> >>
> >> I might be missing something, but
> >> vmlinux provides info of builtin modules
> >> in /sys/module/.
> >
> >No. There are definitely not all modules. I have a builtin sha256_generic,
> >but I can't find him in the /sys/module.
>
> Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> parameters, otherwise you won't find it there. As Masahiro already mentioned,
> if a builtin module has any parameters, they would be accessible under /sys/module/.
>
> >> (Looks like currently only module_param and MODULE_VERSION)
> >>
> >> This patch is not exactly the same, but I see a kind of overwrap.
> >> I'd like to be sure if we want this new scheme.
> >
> >The /sys/module is only for running kernel. One of my use cases is
> >to create an initrd for a new kernel.
> >
> >>
> >> > Information about built-in modules is necessary in the following cases:
> >> >
> >> > 1. When it is necessary to find out what additional parameters can be
> >> > passed to the kernel at boot time.
> >>
> >>
> >> Actually, /sys/module/<module>/parameters/
> >> exposes this information.
> >>
> >> Doesn't it work for your purpose?
> >
> >No, since creating an initrd needs to know all the modalias before
> >I get the sysfs for new kernel. Also there are no modalias at all.
> >
> >> > 2. When you need to know which module names and their aliases are in
> >> > the kernel. This is very useful for creating an initrd image.
> >> >
>
> Hm, I do see one possible additional use-case for preserving module alias
> information for built-in modules - modprobe will currently error (I think,
> correct me if I'm wrong) if we try invoking modprobe with an alias of a
> built-in module, simply because this information is not in modules.builtin or
> modules.alias.

Yes. Patch for modprobe in my todo list. The reason I didn’t do it was
because I wasn’t sure that the file format was final.

> Since kbuild already outputs modules.builtin, I would suggest outputting
> something like a modules.builtin.alias file (and I guess maybe a modules.builtin.param
> file too if that's deemed useful), in a format that is consumable by kmod/modprobe,
> so that modprobing an alias of a built-in module doesn't produce an error. I
> think this should be easy to do if we keep and parse the resulting .modinfo for
> builtin modules. This is just an idea, opinions welcome. I've added Lucas to CC
> in case he has any thoughts.

You don't like kernel.builtin.modinfo ?

It is much easier to create and it has almost the same format as the
modules. So I think it will be easier to parse in kmod.

--
Rgrds, legion


2019-03-28 17:44:22

by Lucas De Marchi

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Wed, Mar 27, 2019 at 8:40 AM Jessica Yu <[email protected]> wrote:
> >No. There are definitely not all modules. I have a builtin sha256_generic,
> >but I can't find him in the /sys/module.
>
> Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> parameters, otherwise you won't find it there. As Masahiro already mentioned,
> if a builtin module has any parameters, they would be accessible under /sys/module/.

Could we please change that and add the sysfs entry regardless of
what's being discussed
here? Not having the entry there simply because we don't have
parameters for that module
always annoyed me.

Lucas De Marchi

2019-03-28 17:58:21

by Lucas De Marchi

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Wed, Mar 27, 2019 at 9:04 AM Alexey Gladkov <[email protected]> wrote:
>
> On Wed, Mar 27, 2019 at 04:40:25PM +0100, Jessica Yu wrote:
> > +++ Alexey Gladkov [26/03/19 18:24 +0100]:
> > >On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> > >> Hi.
> > >>
> > >> (added some people to CC)
> >
> > (Thanks Masahiro for the CC!)
> >
> > >>
> > >> On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> > >> >
> > >> > Problem:
> > >> >
> > >> > When a kernel module is compiled as a separate module, some important
> > >> > information about the kernel module is available via .modinfo section of
> > >> > the module. In contrast, when the kernel module is compiled into the
> > >> > kernel, that information is not available.
> > >>
> > >>
> > >> I might be missing something, but
> > >> vmlinux provides info of builtin modules
> > >> in /sys/module/.
> > >
> > >No. There are definitely not all modules. I have a builtin sha256_generic,
> > >but I can't find him in the /sys/module.
> >
> > Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> > parameters, otherwise you won't find it there. As Masahiro already mentioned,
> > if a builtin module has any parameters, they would be accessible under /sys/module/.
> >
> > >> (Looks like currently only module_param and MODULE_VERSION)
> > >>
> > >> This patch is not exactly the same, but I see a kind of overwrap.
> > >> I'd like to be sure if we want this new scheme.
> > >
> > >The /sys/module is only for running kernel. One of my use cases is
> > >to create an initrd for a new kernel.
> > >
> > >>
> > >> > Information about built-in modules is necessary in the following cases:
> > >> >
> > >> > 1. When it is necessary to find out what additional parameters can be
> > >> > passed to the kernel at boot time.
> > >>
> > >>
> > >> Actually, /sys/module/<module>/parameters/
> > >> exposes this information.
> > >>
> > >> Doesn't it work for your purpose?
> > >
> > >No, since creating an initrd needs to know all the modalias before
> > >I get the sysfs for new kernel. Also there are no modalias at all.
> > >
> > >> > 2. When you need to know which module names and their aliases are in
> > >> > the kernel. This is very useful for creating an initrd image.
> > >> >
> >
> > Hm, I do see one possible additional use-case for preserving module alias
> > information for built-in modules - modprobe will currently error (I think,
> > correct me if I'm wrong) if we try invoking modprobe with an alias of a
> > built-in module, simply because this information is not in modules.builtin or
> > modules.alias.
>
> Yes. Patch for modprobe in my todo list. The reason I didn’t do it was
> because I wasn’t sure that the file format was final.
>
> > Since kbuild already outputs modules.builtin, I would suggest outputting
> > something like a modules.builtin.alias file (and I guess maybe a modules.builtin.param
> > file too if that's deemed useful), in a format that is consumable by kmod/modprobe,
> > so that modprobing an alias of a built-in module doesn't produce an error. I
> > think this should be easy to do if we keep and parse the resulting .modinfo for
> > builtin modules. This is just an idea, opinions welcome. I've added Lucas to CC
> > in case he has any thoughts.
>
> You don't like kernel.builtin.modinfo ?
>
> It is much easier to create and it has almost the same format as the
> modules. So I think it will be easier to parse in kmod.

adding a modules.builtin.alias with the same format of modules.alias
means during modprobe we would only need to load one more file to
lookup aliases. That doesn't mean the kernel built system should do it
though. The same way it's depmod job to create the
modules.alias{,.bin}, we could leave this to depmod if it's in fact
useful to split the information.

I think your version is simple enough and we would get more
information that would be useful for modinfo.
It would indeed be nice to output something useful in "modinfo ext4".

In kmod, I think we would create modules.builtin.modinfo{,.bin} and
just add the aliases to modules.alias{,.bin},
This would keep the names consistent with what is already there.

thanks
Lucas De Marchi

>
> --
> Rgrds, legion
>

2019-03-28 18:46:46

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Thu, Mar 28, 2019 at 10:41:59AM -0700, Lucas De Marchi wrote:
> On Wed, Mar 27, 2019 at 8:40 AM Jessica Yu <[email protected]> wrote:
> > >No. There are definitely not all modules. I have a builtin sha256_generic,
> > >but I can't find him in the /sys/module.
> >
> > Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> > parameters, otherwise you won't find it there. As Masahiro already mentioned,
> > if a builtin module has any parameters, they would be accessible under /sys/module/.
>
> Could we please change that and add the sysfs entry regardless of
> what's being discussed here? Not having the entry there simply because
> we don't have parameters for that module always annoyed me.

What is the sysfs directory going to show? Will it just be empty?

Feel free to send a patch for this, but from what I remember, it wasn't
the easiest thing to do for some reason. But given that the code was
implemented before git was, I can't quite remember.

thanks,

greg k-h

2019-03-28 21:35:14

by Dmitry Torokhov

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Thu, Mar 28, 2019 at 11:45 AM Greg KH <[email protected]> wrote:
>
> On Thu, Mar 28, 2019 at 10:41:59AM -0700, Lucas De Marchi wrote:
> > On Wed, Mar 27, 2019 at 8:40 AM Jessica Yu <[email protected]> wrote:
> > > >No. There are definitely not all modules. I have a builtin sha256_generic,
> > > >but I can't find him in the /sys/module.
> > >
> > > Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> > > parameters, otherwise you won't find it there. As Masahiro already mentioned,
> > > if a builtin module has any parameters, they would be accessible under /sys/module/.
> >
> > Could we please change that and add the sysfs entry regardless of
> > what's being discussed here? Not having the entry there simply because
> > we don't have parameters for that module always annoyed me.

What is the benefit compared to wasting some memory for the directory?

>
> What is the sysfs directory going to show? Will it just be empty?
>
> Feel free to send a patch for this, but from what I remember, it wasn't
> the easiest thing to do for some reason. But given that the code was
> implemented before git was, I can't quite remember.

I am pretty sure we allow empty attribute groups, so it is probbaly as
simple as removing "If (!params) return 0;" form
module_param_sysfs_setup() and making sure we always create
"parameters" group instead of doing it on first visible parameter in
add_sysfs_param(). I suppose we no longer want to skip over parameters
that are not readable nor writable either?

Thanks.

--
Dmitry

2019-04-03 10:47:45

by Masahiro Yamada

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

Hi.

Sorry for the delay.
I added some comments from the
build system point of view.


On Wed, Mar 27, 2019 at 2:24 AM Alexey Gladkov <[email protected]> wrote:
>
> On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> > Hi.
> >
> > (added some people to CC)
> >
> >
> > On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> > >
> > > Problem:
> > >
> > > When a kernel module is compiled as a separate module, some important
> > > information about the kernel module is available via .modinfo section of
> > > the module. In contrast, when the kernel module is compiled into the
> > > kernel, that information is not available.
> >
> >
> > I might be missing something, but
> > vmlinux provides info of builtin modules
> > in /sys/module/.
>
> No. There are definitely not all modules. I have a builtin sha256_generic,
> but I can't find him in the /sys/module.
>
> > (Looks like currently only module_param and MODULE_VERSION)
> >
> > This patch is not exactly the same, but I see a kind of overwrap.
> > I'd like to be sure if we want this new scheme.
>
> The /sys/module is only for running kernel. One of my use cases is
> to create an initrd for a new kernel.
>
> >
> > > Information about built-in modules is necessary in the following cases:
> > >
> > > 1. When it is necessary to find out what additional parameters can be
> > > passed to the kernel at boot time.
> >
> >
> > Actually, /sys/module/<module>/parameters/
> > exposes this information.
> >
> > Doesn't it work for your purpose?
>
> No, since creating an initrd needs to know all the modalias before
> I get the sysfs for new kernel. Also there are no modalias at all.
>
> > > 2. When you need to know which module names and their aliases are in
> > > the kernel. This is very useful for creating an initrd image.
> > >
> > > Proposal:
> > >
> > > The proposed patch does not remove .modinfo section with module
> > > information from the vmlinux at the build time and saves it into a
> > > separate file after kernel linking. So, the kernel does not increase in
> > > size and no additional information remains in it. Information is stored
> > > in the same format as in the separate modules (null-terminated string
> > > array). Because the .modinfo section is already exported with a separate
> > > modules, we are not creating a new API.
> > >
> > > It can be easily read in the userspace:
> > >
> > > $ tr '\0' '\n' < kernel.builtin.modinfo
> > > ext4.softdep=pre: crc32c
> > > ext4.license=GPL
> > > ext4.description=Fourth Extended Filesystem
> > > ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
> > > ext4.alias=fs-ext4
> > > ext4.alias=ext3
> > > ext4.alias=fs-ext3
> > > ext4.alias=ext2
> > > ext4.alias=fs-ext2
> > > md_mod.alias=block-major-9-*
> > > md_mod.alias=md
> > > md_mod.description=MD RAID framework
> > > md_mod.license=GPL
> > > md_mod.parmtype=create_on_open:bool
> > > md_mod.parmtype=start_dirty_degraded:int
> > > ...
> > >
> > > Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > > Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > > Signed-off-by: Alexey Gladkov <[email protected]>
> > > ---
> > > Makefile | 1 +
> > > include/linux/moduleparam.h | 12 +++++-------
> > > scripts/link-vmlinux.sh | 8 ++++++++
> > > 3 files changed, 14 insertions(+), 7 deletions(-)
> > >
> > > diff --git a/Makefile b/Makefile
> > > index d5713e7b1e50..971102194c92 100644
> > > --- a/Makefile
> > > +++ b/Makefile
> > > @@ -1288,6 +1288,7 @@ _modinst_:
> > > fi
> > > @cp -f $(objtree)/modules.order $(MODLIB)/
> > > @cp -f $(objtree)/modules.builtin $(MODLIB)/
> > > + @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/


This file should be cleaned up by 'make clean'.

Please add CLEAN_FILES around line 1336.


# Directories & files removed with 'make clean'
CLEAN_DIRS += $(MODVERDIR) include/ksym
CLEAN_FILES += ...




Also, shouldn't it be added to .gitignore?




> > > $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
> > >
> > > # This depmod is only for convenience to give the initial
> > > diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
> > > index ba36506db4fb..5ba250d9172a 100644
> > > --- a/include/linux/moduleparam.h
> > > +++ b/include/linux/moduleparam.h
> > > @@ -10,23 +10,21 @@
> > > module name. */
> > > #ifdef MODULE
> > > #define MODULE_PARAM_PREFIX /* empty */
> > > +#define __MODULE_INFO_PREFIX /* empty */
> > > #else
> > > #define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
> > > +/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
> > > +#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
> > > #endif
> > >
> > > /* Chosen so that structs with an unsigned long line up. */
> > > #define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
> > >
> > > -#ifdef MODULE
> > > #define __MODULE_INFO(tag, name, info) \
> > > static const char __UNIQUE_ID(name)[] \
> > > __used __attribute__((section(".modinfo"), unused, aligned(1))) \
> > > - = __stringify(tag) "=" info
> > > -#else /* !MODULE */
> > > -/* This struct is here for syntactic coherency, it is not used */
> > > -#define __MODULE_INFO(tag, name, info) \
> > > - struct __UNIQUE_ID(name) {}
> > > -#endif
> > > + = __MODULE_INFO_PREFIX __stringify(tag) "=" info
> > > +
> > > #define __MODULE_PARM_TYPE(name, _type) \
> > > __MODULE_INFO(parmtype, name##type, #name ":" _type)
> > >
> > > diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> > > index c8cf45362bd6..399d7e4d11ec 100755
> > > --- a/scripts/link-vmlinux.sh
> > > +++ b/scripts/link-vmlinux.sh
> > > @@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> > >
> > > # step 1
> > > vmlinux_link "" .tmp_vmlinux1
> > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
> > > kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o
> > >
> > > # step 2
> > > vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
> > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
> > > kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o
> > >
> > > # step 3
> > > @@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> > > kallsyms_vmlinux=.tmp_vmlinux3
> > >
> > > vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
> > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux3
> > >
> > > kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
> > > fi
> > > @@ -281,6 +284,11 @@ fi
> > > info LD vmlinux
> > > vmlinux_link "${kallsymso}" vmlinux
> > >
> > > +info MODINFO kernel.builtin.modinfo
> > > +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
> > > +chmod 444 kernel.builtin.modinfo
> > > +"${OBJCOPY}" -R .modinfo vmlinux
> > > +
> > > if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
> > > info SORTEX vmlinux
> > > sortextable vmlinux



As I pointed out before, I see lots of warnings for powerpc.

How about extracting modinfo before linking vmlinux?
You can extract it from vmlinux.o instead,
and discard it in the linker script.


diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
index dc0e8c5..994cd47 100755
--- a/scripts/link-vmlinux.sh
+++ b/scripts/link-vmlinux.sh
@@ -193,6 +193,9 @@ modpost_link vmlinux.o
# modpost vmlinux.o to check for section mismatches
${MAKE} -f "${srctree}/scripts/Makefile.modpost" vmlinux.o

+info MODINFO kernel.builtin.modinfo
+"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
+
kallsymso=""
kallsyms_vmlinux=""
diff --git a/include/asm-generic/vmlinux.lds.h
b/include/asm-generic/vmlinux.lds.h
index f8f6f04..bbb9e33 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -844,6 +844,7 @@
EXIT_CALL \
*(.discard) \
*(.discard.*) \
+ *(.modinfo) \
}

/**



The code is simpler and I see no more warnings.


I also want you to check MODULE_VERSION.

If CONFIG_SYSFS is enabled, MODULE_VERSION
will not be contained in the file.



Thanks.


> > > --
> > > 2.19.2
> > >
> >
> >
> > --
> > Best Regards
> > Masahiro Yamada
> >
>
> --
> Rgrds, legion
>


--
Best Regards
Masahiro Yamada

2019-04-03 10:50:22

by Masahiro Yamada

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Wed, Apr 3, 2019 at 7:45 PM Masahiro Yamada
<[email protected]> wrote:
>
> Hi.
>
> Sorry for the delay.
> I added some comments from the
> build system point of view.
>
>
> On Wed, Mar 27, 2019 at 2:24 AM Alexey Gladkov <[email protected]> wrote:
> >
> > On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> > > Hi.
> > >
> > > (added some people to CC)
> > >
> > >
> > > On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> > > >
> > > > Problem:
> > > >
> > > > When a kernel module is compiled as a separate module, some important
> > > > information about the kernel module is available via .modinfo section of
> > > > the module. In contrast, when the kernel module is compiled into the
> > > > kernel, that information is not available.
> > >
> > >
> > > I might be missing something, but
> > > vmlinux provides info of builtin modules
> > > in /sys/module/.
> >
> > No. There are definitely not all modules. I have a builtin sha256_generic,
> > but I can't find him in the /sys/module.
> >
> > > (Looks like currently only module_param and MODULE_VERSION)
> > >
> > > This patch is not exactly the same, but I see a kind of overwrap.
> > > I'd like to be sure if we want this new scheme.
> >
> > The /sys/module is only for running kernel. One of my use cases is
> > to create an initrd for a new kernel.
> >
> > >
> > > > Information about built-in modules is necessary in the following cases:
> > > >
> > > > 1. When it is necessary to find out what additional parameters can be
> > > > passed to the kernel at boot time.
> > >
> > >
> > > Actually, /sys/module/<module>/parameters/
> > > exposes this information.
> > >
> > > Doesn't it work for your purpose?
> >
> > No, since creating an initrd needs to know all the modalias before
> > I get the sysfs for new kernel. Also there are no modalias at all.
> >
> > > > 2. When you need to know which module names and their aliases are in
> > > > the kernel. This is very useful for creating an initrd image.
> > > >
> > > > Proposal:
> > > >
> > > > The proposed patch does not remove .modinfo section with module
> > > > information from the vmlinux at the build time and saves it into a
> > > > separate file after kernel linking. So, the kernel does not increase in
> > > > size and no additional information remains in it. Information is stored
> > > > in the same format as in the separate modules (null-terminated string
> > > > array). Because the .modinfo section is already exported with a separate
> > > > modules, we are not creating a new API.
> > > >
> > > > It can be easily read in the userspace:
> > > >
> > > > $ tr '\0' '\n' < kernel.builtin.modinfo
> > > > ext4.softdep=pre: crc32c
> > > > ext4.license=GPL
> > > > ext4.description=Fourth Extended Filesystem
> > > > ext4.author=Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others
> > > > ext4.alias=fs-ext4
> > > > ext4.alias=ext3
> > > > ext4.alias=fs-ext3
> > > > ext4.alias=ext2
> > > > ext4.alias=fs-ext2
> > > > md_mod.alias=block-major-9-*
> > > > md_mod.alias=md
> > > > md_mod.description=MD RAID framework
> > > > md_mod.license=GPL
> > > > md_mod.parmtype=create_on_open:bool
> > > > md_mod.parmtype=start_dirty_degraded:int
> > > > ...
> > > >
> > > > Co-Developed-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > > > Signed-off-by: Gleb Fotengauer-Malinovskiy <[email protected]>
> > > > Signed-off-by: Alexey Gladkov <[email protected]>
> > > > ---
> > > > Makefile | 1 +
> > > > include/linux/moduleparam.h | 12 +++++-------
> > > > scripts/link-vmlinux.sh | 8 ++++++++
> > > > 3 files changed, 14 insertions(+), 7 deletions(-)
> > > >
> > > > diff --git a/Makefile b/Makefile
> > > > index d5713e7b1e50..971102194c92 100644
> > > > --- a/Makefile
> > > > +++ b/Makefile
> > > > @@ -1288,6 +1288,7 @@ _modinst_:
> > > > fi
> > > > @cp -f $(objtree)/modules.order $(MODLIB)/
> > > > @cp -f $(objtree)/modules.builtin $(MODLIB)/
> > > > + @cp -f $(objtree)/kernel.builtin.modinfo $(MODLIB)/
>
>
> This file should be cleaned up by 'make clean'.
>
> Please add CLEAN_FILES around line 1336.
>
>
> # Directories & files removed with 'make clean'
> CLEAN_DIRS += $(MODVERDIR) include/ksym
> CLEAN_FILES += ...
>
>
>
>
> Also, shouldn't it be added to .gitignore?
>
>
>
>
> > > > $(Q)$(MAKE) -f $(srctree)/scripts/Makefile.modinst
> > > >
> > > > # This depmod is only for convenience to give the initial
> > > > diff --git a/include/linux/moduleparam.h b/include/linux/moduleparam.h
> > > > index ba36506db4fb..5ba250d9172a 100644
> > > > --- a/include/linux/moduleparam.h
> > > > +++ b/include/linux/moduleparam.h
> > > > @@ -10,23 +10,21 @@
> > > > module name. */
> > > > #ifdef MODULE
> > > > #define MODULE_PARAM_PREFIX /* empty */
> > > > +#define __MODULE_INFO_PREFIX /* empty */
> > > > #else
> > > > #define MODULE_PARAM_PREFIX KBUILD_MODNAME "."
> > > > +/* We cannot use MODULE_PARAM_PREFIX because some modules override it. */
> > > > +#define __MODULE_INFO_PREFIX KBUILD_MODNAME "."
> > > > #endif
> > > >
> > > > /* Chosen so that structs with an unsigned long line up. */
> > > > #define MAX_PARAM_PREFIX_LEN (64 - sizeof(unsigned long))
> > > >
> > > > -#ifdef MODULE
> > > > #define __MODULE_INFO(tag, name, info) \
> > > > static const char __UNIQUE_ID(name)[] \
> > > > __used __attribute__((section(".modinfo"), unused, aligned(1))) \
> > > > - = __stringify(tag) "=" info
> > > > -#else /* !MODULE */
> > > > -/* This struct is here for syntactic coherency, it is not used */
> > > > -#define __MODULE_INFO(tag, name, info) \
> > > > - struct __UNIQUE_ID(name) {}
> > > > -#endif
> > > > + = __MODULE_INFO_PREFIX __stringify(tag) "=" info
> > > > +
> > > > #define __MODULE_PARM_TYPE(name, _type) \
> > > > __MODULE_INFO(parmtype, name##type, #name ":" _type)
> > > >
> > > > diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> > > > index c8cf45362bd6..399d7e4d11ec 100755
> > > > --- a/scripts/link-vmlinux.sh
> > > > +++ b/scripts/link-vmlinux.sh
> > > > @@ -258,10 +258,12 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> > > >
> > > > # step 1
> > > > vmlinux_link "" .tmp_vmlinux1
> > > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux1
> > > > kallsyms .tmp_vmlinux1 .tmp_kallsyms1.o
> > > >
> > > > # step 2
> > > > vmlinux_link .tmp_kallsyms1.o .tmp_vmlinux2
> > > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux2
> > > > kallsyms .tmp_vmlinux2 .tmp_kallsyms2.o
> > > >
> > > > # step 3
> > > > @@ -273,6 +275,7 @@ if [ -n "${CONFIG_KALLSYMS}" ]; then
> > > > kallsyms_vmlinux=.tmp_vmlinux3
> > > >
> > > > vmlinux_link .tmp_kallsyms2.o .tmp_vmlinux3
> > > > + "${OBJCOPY}" -R .modinfo .tmp_vmlinux3
> > > >
> > > > kallsyms .tmp_vmlinux3 .tmp_kallsyms3.o
> > > > fi
> > > > @@ -281,6 +284,11 @@ fi
> > > > info LD vmlinux
> > > > vmlinux_link "${kallsymso}" vmlinux
> > > >
> > > > +info MODINFO kernel.builtin.modinfo
> > > > +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo
> > > > +chmod 444 kernel.builtin.modinfo
> > > > +"${OBJCOPY}" -R .modinfo vmlinux
> > > > +
> > > > if [ -n "${CONFIG_BUILDTIME_EXTABLE_SORT}" ]; then
> > > > info SORTEX vmlinux
> > > > sortextable vmlinux
>
>
>
> As I pointed out before, I see lots of warnings for powerpc.
>
> How about extracting modinfo before linking vmlinux?
> You can extract it from vmlinux.o instead,
> and discard it in the linker script.
>
>
> diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> index dc0e8c5..994cd47 100755
> --- a/scripts/link-vmlinux.sh
> +++ b/scripts/link-vmlinux.sh
> @@ -193,6 +193,9 @@ modpost_link vmlinux.o
> # modpost vmlinux.o to check for section mismatches
> ${MAKE} -f "${srctree}/scripts/Makefile.modpost" vmlinux.o
>
> +info MODINFO kernel.builtin.modinfo
> +"${OBJCOPY}" -j .modinfo -O binary vmlinux kernel.builtin.modinfo

Oops, this should be:

"${OBJCOPY}" -j .modinfo -O binary vmlinux.o kernel.builtin.modinfo




--
Best Regards
Masahiro Yamada

2019-04-03 11:33:02

by Masahiro Yamada

[permalink] [raw]
Subject: Re: [RESEND PATCH v1] moduleparam: Save information about built-in modules in separate file

On Thu, Mar 28, 2019 at 1:04 AM Alexey Gladkov <[email protected]> wrote:
>
> On Wed, Mar 27, 2019 at 04:40:25PM +0100, Jessica Yu wrote:
> > +++ Alexey Gladkov [26/03/19 18:24 +0100]:
> > >On Fri, Mar 22, 2019 at 02:34:12PM +0900, Masahiro Yamada wrote:
> > >> Hi.
> > >>
> > >> (added some people to CC)
> >
> > (Thanks Masahiro for the CC!)
> >
> > >>
> > >> On Fri, Mar 15, 2019 at 7:10 PM Alexey Gladkov <[email protected]> wrote:
> > >> >
> > >> > Problem:
> > >> >
> > >> > When a kernel module is compiled as a separate module, some important
> > >> > information about the kernel module is available via .modinfo section of
> > >> > the module. In contrast, when the kernel module is compiled into the
> > >> > kernel, that information is not available.
> > >>
> > >>
> > >> I might be missing something, but
> > >> vmlinux provides info of builtin modules
> > >> in /sys/module/.
> > >
> > >No. There are definitely not all modules. I have a builtin sha256_generic,
> > >but I can't find him in the /sys/module.
> >
> > Yeah, you'll only find builtin modules under /sys/module/ if it has any module
> > parameters, otherwise you won't find it there. As Masahiro already mentioned,
> > if a builtin module has any parameters, they would be accessible under /sys/module/.
> >
> > >> (Looks like currently only module_param and MODULE_VERSION)
> > >>
> > >> This patch is not exactly the same, but I see a kind of overwrap.
> > >> I'd like to be sure if we want this new scheme.
> > >
> > >The /sys/module is only for running kernel. One of my use cases is
> > >to create an initrd for a new kernel.
> > >
> > >>
> > >> > Information about built-in modules is necessary in the following cases:
> > >> >
> > >> > 1. When it is necessary to find out what additional parameters can be
> > >> > passed to the kernel at boot time.
> > >>
> > >>
> > >> Actually, /sys/module/<module>/parameters/
> > >> exposes this information.
> > >>
> > >> Doesn't it work for your purpose?
> > >
> > >No, since creating an initrd needs to know all the modalias before
> > >I get the sysfs for new kernel. Also there are no modalias at all.
> > >
> > >> > 2. When you need to know which module names and their aliases are in
> > >> > the kernel. This is very useful for creating an initrd image.
> > >> >
> >
> > Hm, I do see one possible additional use-case for preserving module alias
> > information for built-in modules - modprobe will currently error (I think,
> > correct me if I'm wrong) if we try invoking modprobe with an alias of a
> > built-in module, simply because this information is not in modules.builtin or
> > modules.alias.
>
> Yes. Patch for modprobe in my todo list. The reason I didn’t do it was
> because I wasn’t sure that the file format was final.
>
> > Since kbuild already outputs modules.builtin, I would suggest outputting
> > something like a modules.builtin.alias file (and I guess maybe a modules.builtin.param
> > file too if that's deemed useful), in a format that is consumable by kmod/modprobe,
> > so that modprobing an alias of a built-in module doesn't produce an error. I
> > think this should be easy to do if we keep and parse the resulting .modinfo for
> > builtin modules. This is just an idea, opinions welcome. I've added Lucas to CC
> > in case he has any thoughts.
>
> You don't like kernel.builtin.modinfo ?


Naming is often the most difficult thing. :)

IMHO, 'kernel' and 'builtin' have a similar meaning here.
Is 'kernel.builtin' unnecessarily too long?


Perhaps, another idea is:

'builtin.alias' instead of 'modules.builtin.alias'
if we want a separate file in the same format.


In hindsight, 'modules.builtin' should have been
'builtin.order', I think.




> It is much easier to create and it has almost the same format as the
> modules. So I think it will be easier to parse in kmod.

--
Best Regards
Masahiro Yamada