2022-05-19 20:01:12

by Naveen N. Rao

[permalink] [raw]
Subject: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]

Since commit d1bcae833b32f1 ("ELF: Don't generate unused section
symbols") [1], binutils (v2.36+) started dropping section symbols that
it thought were unused. This isn't an issue in general, but with
kexec_file.c, gcc is placing kexec_arch_apply_relocations[_add] into a
separate .text.unlikely section and the section symbol ".text.unlikely"
is being dropped. Due to this, recordmcount is unable to find a non-weak
symbol in .text.unlikely to generate a relocation record against.

Address this by dropping the weak attribute from these functions.
Instead, follow the existing pattern of having architectures #define the
name of the function they want to override in their headers.

[1] https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=d1bcae833b32f1

Signed-off-by: Michael Ellerman <[email protected]>
Signed-off-by: Naveen N. Rao <[email protected]>
---
arch/s390/include/asm/kexec.h | 8 ++++++
arch/x86/include/asm/kexec.h | 8 ++++++
include/linux/kexec.h | 46 +++++++++++++++++++++++++++++------
kernel/kexec_file.c | 34 --------------------------
4 files changed, 54 insertions(+), 42 deletions(-)

diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h
index 7f3c9ac34bd8d1..540dd469b0886d 100644
--- a/arch/s390/include/asm/kexec.h
+++ b/arch/s390/include/asm/kexec.h
@@ -83,4 +83,12 @@ struct kimage_arch {
extern const struct kexec_file_ops s390_kexec_image_ops;
extern const struct kexec_file_ops s390_kexec_elf_ops;

+#ifdef CONFIG_KEXEC_FILE
+struct purgatory_info;
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+ Elf_Shdr *section,
+ const Elf_Shdr *relsec,
+ const Elf_Shdr *symtab);
+#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
+#endif
#endif /*_S390_KEXEC_H */
diff --git a/arch/x86/include/asm/kexec.h b/arch/x86/include/asm/kexec.h
index 11b7c06e2828c3..6ad8d946cd3ebf 100644
--- a/arch/x86/include/asm/kexec.h
+++ b/arch/x86/include/asm/kexec.h
@@ -186,6 +186,14 @@ extern int arch_kexec_post_alloc_pages(void *vaddr, unsigned int pages,
extern void arch_kexec_pre_free_pages(void *vaddr, unsigned int pages);
#define arch_kexec_pre_free_pages arch_kexec_pre_free_pages

+#ifdef CONFIG_KEXEC_FILE
+struct purgatory_info;
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+ Elf_Shdr *section,
+ const Elf_Shdr *relsec,
+ const Elf_Shdr *symtab);
+#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
+#endif
#endif

typedef void crash_vmclear_fn(void);
diff --git a/include/linux/kexec.h b/include/linux/kexec.h
index 58d1b58a971e34..fcd5035209f19a 100644
--- a/include/linux/kexec.h
+++ b/include/linux/kexec.h
@@ -193,14 +193,6 @@ void *kexec_purgatory_get_symbol_addr(struct kimage *image, const char *name);
int arch_kexec_kernel_image_probe(struct kimage *image, void *buf,
unsigned long buf_len);
void *arch_kexec_kernel_image_load(struct kimage *image);
-int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
- Elf_Shdr *section,
- const Elf_Shdr *relsec,
- const Elf_Shdr *symtab);
-int arch_kexec_apply_relocations(struct purgatory_info *pi,
- Elf_Shdr *section,
- const Elf_Shdr *relsec,
- const Elf_Shdr *symtab);
int arch_kimage_file_post_load_cleanup(struct kimage *image);
#ifdef CONFIG_KEXEC_SIG
int arch_kexec_kernel_verify_sig(struct kimage *image, void *buf,
@@ -229,6 +221,44 @@ extern int crash_exclude_mem_range(struct crash_mem *mem,
unsigned long long mend);
extern int crash_prepare_elf64_headers(struct crash_mem *mem, int kernel_map,
void **addr, unsigned long *sz);
+
+#ifndef arch_kexec_apply_relocations_add
+/*
+ * arch_kexec_apply_relocations_add - apply relocations of type RELA
+ * @pi: Purgatory to be relocated.
+ * @section: Section relocations applying to.
+ * @relsec: Section containing RELAs.
+ * @symtab: Corresponding symtab.
+ *
+ * Return: 0 on success, negative errno on error.
+ */
+static inline int
+arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
+ const Elf_Shdr *relsec, const Elf_Shdr *symtab)
+{
+ pr_err("RELA relocation unsupported.\n");
+ return -ENOEXEC;
+}
+#endif
+
+#ifndef arch_kexec_apply_relocations
+/*
+ * arch_kexec_apply_relocations - apply relocations of type REL
+ * @pi: Purgatory to be relocated.
+ * @section: Section relocations applying to.
+ * @relsec: Section containing RELs.
+ * @symtab: Corresponding symtab.
+ *
+ * Return: 0 on success, negative errno on error.
+ */
+static inline int
+arch_kexec_apply_relocations(struct purgatory_info *pi, Elf_Shdr *section,
+ const Elf_Shdr *relsec, const Elf_Shdr *symtab)
+{
+ pr_err("REL relocation unsupported.\n");
+ return -ENOEXEC;
+}
+#endif
#endif /* CONFIG_KEXEC_FILE */

#ifdef CONFIG_KEXEC_ELF
diff --git a/kernel/kexec_file.c b/kernel/kexec_file.c
index 8347fc158d2b96..c108a2a8875463 100644
--- a/kernel/kexec_file.c
+++ b/kernel/kexec_file.c
@@ -108,40 +108,6 @@ int __weak arch_kexec_kernel_verify_sig(struct kimage *image, void *buf,
}
#endif

-/*
- * arch_kexec_apply_relocations_add - apply relocations of type RELA
- * @pi: Purgatory to be relocated.
- * @section: Section relocations applying to.
- * @relsec: Section containing RELAs.
- * @symtab: Corresponding symtab.
- *
- * Return: 0 on success, negative errno on error.
- */
-int __weak
-arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
- const Elf_Shdr *relsec, const Elf_Shdr *symtab)
-{
- pr_err("RELA relocation unsupported.\n");
- return -ENOEXEC;
-}
-
-/*
- * arch_kexec_apply_relocations - apply relocations of type REL
- * @pi: Purgatory to be relocated.
- * @section: Section relocations applying to.
- * @relsec: Section containing RELs.
- * @symtab: Corresponding symtab.
- *
- * Return: 0 on success, negative errno on error.
- */
-int __weak
-arch_kexec_apply_relocations(struct purgatory_info *pi, Elf_Shdr *section,
- const Elf_Shdr *relsec, const Elf_Shdr *symtab)
-{
- pr_err("REL relocation unsupported.\n");
- return -ENOEXEC;
-}
-
/*
* Free up memory used by kernel, initrd, and command line. This is temporary
* memory allocation which is not needed any more after these buffers have

base-commit: f993aed406eaf968ba3867a76bb46c95336a33d0
--
2.36.1



2022-05-21 07:00:10

by kernel test robot

[permalink] [raw]
Subject: Re: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]

Hi "Naveen,

I love your patch! Yet something to improve:

[auto build test ERROR on f993aed406eaf968ba3867a76bb46c95336a33d0]

url: https://github.com/intel-lab-lkp/linux/commits/Naveen-N-Rao/kexec_file-Drop-weak-attribute-from-arch_kexec_apply_relocations-_add/20220519-171432
base: f993aed406eaf968ba3867a76bb46c95336a33d0
config: s390-allmodconfig (https://download.01.org/0day-ci/archive/20220519/[email protected]/config)
compiler: s390-linux-gcc (GCC) 11.3.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/intel-lab-lkp/linux/commit/67171688c71cb5b05f26e0dfc45eec8d8d1428ff
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review Naveen-N-Rao/kexec_file-Drop-weak-attribute-from-arch_kexec_apply_relocations-_add/20220519-171432
git checkout 67171688c71cb5b05f26e0dfc45eec8d8d1428ff
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.3.0 make.cross W=1 O=build_dir ARCH=s390 SHELL=/bin/bash arch/s390/

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <[email protected]>

All errors (new ones prefixed by >>):

In file included from arch/s390/kernel/machine_kexec_reloc.c:3:
>> arch/s390/include/asm/kexec.h:89:38: error: unknown type name 'Elf_Shdr'; did you mean 'elf_shdr'?
89 | Elf_Shdr *section,
| ^~~~~~~~
| elf_shdr
arch/s390/include/asm/kexec.h:90:44: error: unknown type name 'Elf_Shdr'
90 | const Elf_Shdr *relsec,
| ^~~~~~~~
arch/s390/include/asm/kexec.h:91:44: error: unknown type name 'Elf_Shdr'
91 | const Elf_Shdr *symtab);
| ^~~~~~~~


vim +89 arch/s390/include/asm/kexec.h

85
86 #ifdef CONFIG_KEXEC_FILE
87 struct purgatory_info;
88 int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> 89 Elf_Shdr *section,

--
0-DAY CI Kernel Test Service
https://01.org/lkp

2022-05-29 07:50:20

by Uwe Kleine-König

[permalink] [raw]
Subject: Re: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]

Hello,

on current linux-next ARCH=riscv allmodconfig breaks with:

CC arch/riscv/kernel/elf_kexec.o
arch/riscv/kernel/elf_kexec.c:345:5: error: redefinition of ‘arch_kexec_apply_relocations_add’
345 | int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from arch/riscv/kernel/elf_kexec.c:16:
include/linux/kexec.h:236:1: note: previous definition of ‘arch_kexec_apply_relocations_add’ with type ‘int(struct purgatory_info *, Elf64_Shdr *, const Elf64_Shdr *, const Elf64_Shdr *)’ {aka ‘int(struct purgatory_info *, struct elf64_shdr *, const struct elf64_shdr *, const struct elf64_shdr *)’}
236 | arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

(I think) because there is a conflict between the two commits:

233c1e6c319c kexec_file: drop weak attribute from arch_kexec_apply_relocations[_add]
838b3e28488f RISC-V: Load purgatory in kexec_file

And so next is broken starting from
164a9037b1d33f28ba27671c16ec1c23d4a11acf which merges the riscv tree.

Best regards
Uwe

--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (1.30 kB)
signature.asc (499.00 B)
Download all attachments

2022-05-29 11:16:28

by Christophe Leroy

[permalink] [raw]
Subject: Re: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]



Le 29/05/2022 à 08:56, Uwe Kleine-König a écrit :
> Hello,
>
> on current linux-next ARCH=riscv allmodconfig breaks with:
>
> CC arch/riscv/kernel/elf_kexec.o
> arch/riscv/kernel/elf_kexec.c:345:5: error: redefinition of ‘arch_kexec_apply_relocations_add’
> 345 | int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> In file included from arch/riscv/kernel/elf_kexec.c:16:
> include/linux/kexec.h:236:1: note: previous definition of ‘arch_kexec_apply_relocations_add’ with type ‘int(struct purgatory_info *, Elf64_Shdr *, const Elf64_Shdr *, const Elf64_Shdr *)’ {aka ‘int(struct purgatory_info *, struct elf64_shdr *, const struct elf64_shdr *, const struct elf64_shdr *)’}
> 236 | arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> (I think) because there is a conflict between the two commits:
>
> 233c1e6c319c kexec_file: drop weak attribute from arch_kexec_apply_relocations[_add]
> 838b3e28488f RISC-V: Load purgatory in kexec_file
>
> And so next is broken starting from
> 164a9037b1d33f28ba27671c16ec1c23d4a11acf which merges the riscv tree.
>

In arch/riscv/include/asm/kexec.h, do the same as s390 did in commit
233c1e6c319c:


diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h
index 7f3c9ac34bd8..540dd469b088 100644
--- a/arch/s390/include/asm/kexec.h
+++ b/arch/s390/include/asm/kexec.h
@@ -83,4 +83,12 @@ struct kimage_arch {
extern const struct kexec_file_ops s390_kexec_image_ops;
extern const struct kexec_file_ops s390_kexec_elf_ops;

+#ifdef CONFIG_KEXEC_FILE
+struct purgatory_info;
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+ Elf_Shdr *section,
+ const Elf_Shdr *relsec,
+ const Elf_Shdr *symtab);
+#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
+#endif
#endif /*_S390_KEXEC_H */

2022-05-30 02:28:02

by liaochang (A)

[permalink] [raw]
Subject: Re: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]



在 2022/5/29 22:06, Uwe Kleine-König 写道:
> Hello,
>
> On Sun, May 29, 2022 at 09:46:47AM +0000, Christophe Leroy wrote:
>> Le 29/05/2022 à 08:56, Uwe Kleine-König a écrit :
>>> Hello,
>>>
>>> on current linux-next ARCH=riscv allmodconfig breaks with:
>>>
>>> CC arch/riscv/kernel/elf_kexec.o
>>> arch/riscv/kernel/elf_kexec.c:345:5: error: redefinition of ‘arch_kexec_apply_relocations_add’
>>> 345 | int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
>>> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>> In file included from arch/riscv/kernel/elf_kexec.c:16:
>>> include/linux/kexec.h:236:1: note: previous definition of ‘arch_kexec_apply_relocations_add’ with type ‘int(struct purgatory_info *, Elf64_Shdr *, const Elf64_Shdr *, const Elf64_Shdr *)’ {aka ‘int(struct purgatory_info *, struct elf64_shdr *, const struct elf64_shdr *, const struct elf64_shdr *)’}
>>> 236 | arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
>>> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>
>>> (I think) because there is a conflict between the two commits:
>>>
>>> 233c1e6c319c kexec_file: drop weak attribute from arch_kexec_apply_relocations[_add]
>>> 838b3e28488f RISC-V: Load purgatory in kexec_file
>>>
>>> And so next is broken starting from
>>> 164a9037b1d33f28ba27671c16ec1c23d4a11acf which merges the riscv tree.
>>>
>>
>> In arch/riscv/include/asm/kexec.h, do the same as s390 did in commit
>> 233c1e6c319c:
>>
>> diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h
>> index 7f3c9ac34bd8..540dd469b088 100644
>> --- a/arch/s390/include/asm/kexec.h
>> +++ b/arch/s390/include/asm/kexec.h
>> @@ -83,4 +83,12 @@ struct kimage_arch {
>> extern const struct kexec_file_ops s390_kexec_image_ops;
>> extern const struct kexec_file_ops s390_kexec_elf_ops;
>>
>> +#ifdef CONFIG_KEXEC_FILE
>> +struct purgatory_info;
>> +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
>> + Elf_Shdr *section,
>> + const Elf_Shdr *relsec,
>> + const Elf_Shdr *symtab);
>> +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
>> +#endif
>> #endif /*_S390_KEXEC_H */
>
> I can confirm that doing
>
> diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
> index 206217b23301..eee260e8ab30 100644
> --- a/arch/riscv/include/asm/kexec.h
> +++ b/arch/riscv/include/asm/kexec.h
> @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
>
> #ifdef CONFIG_KEXEC_FILE
> extern const struct kexec_file_ops elf_kexec_ops;
> +
> +struct purgatory_info;
> +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> + Elf_Shdr *section,
> + const Elf_Shdr *relsec,
> + const Elf_Shdr *symtab);
> +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> #endif
>
> #endif

LGTM, you could send a fixup patch to riscv, thanks.

>
> on top of 838b3e28488f results in a compilable tree. And when merging
> 233c1e6c319c into this, it is still building.
>
> I'm not enough into kexec (and riscv) to judge if this is sensible, or
> create a useful commit log but the obvious way forward is to apply the
> above patch to the riscv tree before it hits Linus' tree.
>
> Best regards
> Uwe
>
>
> _______________________________________________
> linux-riscv mailing list
> [email protected]
> http://lists.infradead.org/mailman/listinfo/linux-riscv

--
BR,
Liao, Chang

2022-05-30 10:54:43

by Uwe Kleine-König

[permalink] [raw]
Subject: Re: [PATCH v2] kexec_file: Drop weak attribute from arch_kexec_apply_relocations[_add]

Hello,

On Sun, May 29, 2022 at 09:46:47AM +0000, Christophe Leroy wrote:
> Le 29/05/2022 à 08:56, Uwe Kleine-König a écrit :
> > Hello,
> >
> > on current linux-next ARCH=riscv allmodconfig breaks with:
> >
> > CC arch/riscv/kernel/elf_kexec.o
> > arch/riscv/kernel/elf_kexec.c:345:5: error: redefinition of ‘arch_kexec_apply_relocations_add’
> > 345 | int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> > | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > In file included from arch/riscv/kernel/elf_kexec.c:16:
> > include/linux/kexec.h:236:1: note: previous definition of ‘arch_kexec_apply_relocations_add’ with type ‘int(struct purgatory_info *, Elf64_Shdr *, const Elf64_Shdr *, const Elf64_Shdr *)’ {aka ‘int(struct purgatory_info *, struct elf64_shdr *, const struct elf64_shdr *, const struct elf64_shdr *)’}
> > 236 | arch_kexec_apply_relocations_add(struct purgatory_info *pi, Elf_Shdr *section,
> > | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >
> > (I think) because there is a conflict between the two commits:
> >
> > 233c1e6c319c kexec_file: drop weak attribute from arch_kexec_apply_relocations[_add]
> > 838b3e28488f RISC-V: Load purgatory in kexec_file
> >
> > And so next is broken starting from
> > 164a9037b1d33f28ba27671c16ec1c23d4a11acf which merges the riscv tree.
> >
>
> In arch/riscv/include/asm/kexec.h, do the same as s390 did in commit
> 233c1e6c319c:
>
> diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h
> index 7f3c9ac34bd8..540dd469b088 100644
> --- a/arch/s390/include/asm/kexec.h
> +++ b/arch/s390/include/asm/kexec.h
> @@ -83,4 +83,12 @@ struct kimage_arch {
> extern const struct kexec_file_ops s390_kexec_image_ops;
> extern const struct kexec_file_ops s390_kexec_elf_ops;
>
> +#ifdef CONFIG_KEXEC_FILE
> +struct purgatory_info;
> +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> + Elf_Shdr *section,
> + const Elf_Shdr *relsec,
> + const Elf_Shdr *symtab);
> +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> +#endif
> #endif /*_S390_KEXEC_H */

I can confirm that doing

diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
index 206217b23301..eee260e8ab30 100644
--- a/arch/riscv/include/asm/kexec.h
+++ b/arch/riscv/include/asm/kexec.h
@@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;

#ifdef CONFIG_KEXEC_FILE
extern const struct kexec_file_ops elf_kexec_ops;
+
+struct purgatory_info;
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+ Elf_Shdr *section,
+ const Elf_Shdr *relsec,
+ const Elf_Shdr *symtab);
+#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
#endif

#endif

on top of 838b3e28488f results in a compilable tree. And when merging
233c1e6c319c into this, it is still building.

I'm not enough into kexec (and riscv) to judge if this is sensible, or
create a useful commit log but the obvious way forward is to apply the
above patch to the riscv tree before it hits Linus' tree.

Best regards
Uwe

--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (3.29 kB)
signature.asc (499.00 B)
Download all attachments

2022-05-30 14:12:59

by Uwe Kleine-König

[permalink] [raw]
Subject: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
commit 233c1e6c319c ("kexec_file: drop weak attribute from
arch_kexec_apply_relocations[_add]") is also contained in the tree.
This currently happens in next-20220527.

Prepare the RISC-V similar to the s390 adaption done in 233c1e6c319c.
This is safe to do on top of the riscv change even without the change to
arch_kexec_apply_relocations.

Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
Looks-good-to: liaochang (A) <[email protected]>
Signed-off-by: Uwe Kleine-K?nig <[email protected]>
---

On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
> > I can confirm that doing
> >
> > diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
> > index 206217b23301..eee260e8ab30 100644
> > --- a/arch/riscv/include/asm/kexec.h
> > +++ b/arch/riscv/include/asm/kexec.h
> > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
> >
> > #ifdef CONFIG_KEXEC_FILE
> > extern const struct kexec_file_ops elf_kexec_ops;
> > +
> > +struct purgatory_info;
> > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> > + Elf_Shdr *section,
> > + const Elf_Shdr *relsec,
> > + const Elf_Shdr *symtab);
> > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> > #endif
> >
> > #endif
>
> LGTM, you could send a fixup patch to riscv, thanks.
>
> >
> > on top of 838b3e28488f results in a compilable tree. And when merging
> > 233c1e6c319c into this, it is still building.
> >
> > I'm not enough into kexec (and riscv) to judge if this is sensible, or
> > create a useful commit log but the obvious way forward is to apply the
> > above patch to the riscv tree before it hits Linus' tree.

Ok, here comes a patch with a generic commit log.

@riscv people: If you prefer, squash it into 838b3e28488f.

Best regards
Uwe

arch/riscv/include/asm/kexec.h | 7 +++++++
1 file changed, 7 insertions(+)

diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
index 206217b23301..eee260e8ab30 100644
--- a/arch/riscv/include/asm/kexec.h
+++ b/arch/riscv/include/asm/kexec.h
@@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;

#ifdef CONFIG_KEXEC_FILE
extern const struct kexec_file_ops elf_kexec_ops;
+
+struct purgatory_info;
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+ Elf_Shdr *section,
+ const Elf_Shdr *relsec,
+ const Elf_Shdr *symtab);
+#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
#endif

#endif

base-commit: 838b3e28488f702e2b5477b393f009b2639d2b1a
--
2.36.1

--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (2.86 kB)
signature.asc (499.00 B)
Download all attachments

2022-05-31 01:24:41

by Palmer Dabbelt

[permalink] [raw]
Subject: Re: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

On Mon, 30 May 2022 00:42:02 PDT (-0700), [email protected] wrote:
> Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
> commit 233c1e6c319c ("kexec_file: drop weak attribute from
> arch_kexec_apply_relocations[_add]") is also contained in the tree.
> This currently happens in next-20220527.
>
> Prepare the RISC-V similar to the s390 adaption done in 233c1e6c319c.
> This is safe to do on top of the riscv change even without the change to
> arch_kexec_apply_relocations.
>
> Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
> Looks-good-to: liaochang (A) <[email protected]>
> Signed-off-by: Uwe Kleine-König <[email protected]>
> ---
>
> On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
>> > I can confirm that doing
>> >
>> > diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
>> > index 206217b23301..eee260e8ab30 100644
>> > --- a/arch/riscv/include/asm/kexec.h
>> > +++ b/arch/riscv/include/asm/kexec.h
>> > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
>> >
>> > #ifdef CONFIG_KEXEC_FILE
>> > extern const struct kexec_file_ops elf_kexec_ops;
>> > +
>> > +struct purgatory_info;
>> > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
>> > + Elf_Shdr *section,
>> > + const Elf_Shdr *relsec,
>> > + const Elf_Shdr *symtab);
>> > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
>> > #endif
>> >
>> > #endif
>>
>> LGTM, you could send a fixup patch to riscv, thanks.
>>
>> >
>> > on top of 838b3e28488f results in a compilable tree. And when merging
>> > 233c1e6c319c into this, it is still building.
>> >
>> > I'm not enough into kexec (and riscv) to judge if this is sensible, or
>> > create a useful commit log but the obvious way forward is to apply the
>> > above patch to the riscv tree before it hits Linus' tree.
>
> Ok, here comes a patch with a generic commit log.
>
> @riscv people: If you prefer, squash it into 838b3e28488f.

Sorry, just saw this after I sent my version of the fix. They're the
same, but do you mind sending a full-on patch so I can merge it?

>
> Best regards
> Uwe
>
> arch/riscv/include/asm/kexec.h | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/arch/riscv/include/asm/kexec.h b/arch/riscv/include/asm/kexec.h
> index 206217b23301..eee260e8ab30 100644
> --- a/arch/riscv/include/asm/kexec.h
> +++ b/arch/riscv/include/asm/kexec.h
> @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
>
> #ifdef CONFIG_KEXEC_FILE
> extern const struct kexec_file_ops elf_kexec_ops;
> +
> +struct purgatory_info;
> +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> + Elf_Shdr *section,
> + const Elf_Shdr *relsec,
> + const Elf_Shdr *symtab);
> +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> #endif
>
> #endif
>
> base-commit: 838b3e28488f702e2b5477b393f009b2639d2b1a
> --
> 2.36.1
>
> --
> Pengutronix e.K. | Uwe Kleine-König |
> Industrial Linux Solutions | https://www.pengutronix.de/ |

2022-05-31 16:20:14

by Palmer Dabbelt

[permalink] [raw]
Subject: Re: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

On Mon, 30 May 2022 12:41:33 PDT (-0700), [email protected] wrote:
> Hello,
>
> On Mon, May 30, 2022 at 11:58:16AM -0700, Palmer Dabbelt wrote:
>> On Mon, 30 May 2022 00:42:02 PDT (-0700), [email protected] wrote:
>> > Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
>> > commit 233c1e6c319c ("kexec_file: drop weak attribute from
>> > arch_kexec_apply_relocations[_add]") is also contained in the tree.
>> > This currently happens in next-20220527.
>> >
>> > Prepare the RISC-V similar to the s390 adaption done in 233c1e6c319c.
>> > This is safe to do on top of the riscv change even without the change to
>> > arch_kexec_apply_relocations.
>> >
>> > Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
>> > Looks-good-to: liaochang (A) <[email protected]>
>> > Signed-off-by: Uwe Kleine-König <[email protected]>
>> > ---
>> >
>> > On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
>> > > > I can confirm that doing
>> > > > > diff --git a/arch/riscv/include/asm/kexec.h
>> > > b/arch/riscv/include/asm/kexec.h
>> > > > index 206217b23301..eee260e8ab30 100644
>> > > > --- a/arch/riscv/include/asm/kexec.h
>> > > > +++ b/arch/riscv/include/asm/kexec.h
>> > > > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
>> > > > > #ifdef CONFIG_KEXEC_FILE
>> > > > extern const struct kexec_file_ops elf_kexec_ops;
>> > > > +
>> > > > +struct purgatory_info;
>> > > > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
>> > > > + Elf_Shdr *section,
>> > > > + const Elf_Shdr *relsec,
>> > > > + const Elf_Shdr *symtab);
>> > > > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
>> > > > #endif
>> > > > > #endif
>> > >
>> > > LGTM, you could send a fixup patch to riscv, thanks.
>> > >
>> > > > > on top of 838b3e28488f results in a compilable tree. And when
>> > > merging
>> > > > 233c1e6c319c into this, it is still building.
>> > > > > I'm not enough into kexec (and riscv) to judge if this is
>> > > sensible, or
>> > > > create a useful commit log but the obvious way forward is to apply the
>> > > > above patch to the riscv tree before it hits Linus' tree.
>> >
>> > Ok, here comes a patch with a generic commit log.
>> >
>> > @riscv people: If you prefer, squash it into 838b3e28488f.
>>
>> Sorry, just saw this after I sent my version of the fix. They're the same,
>> but do you mind sending a full-on patch so I can merge it?
>
> Sorry, I don't understand your request. I found
> https://lore.kernel.org/linux-riscv/[email protected]/
>
> but I don't know what a full-on patch is and what stops you merging my
> patch.
>
> Is it that it's in reply to a patch series and b4 fails to fetch the
> right patch and you ask to send it in a new thread?

Ya, with the reply bits in there my mail merge (which unfortunately
isn't b4, I haven't gotten around to converting yet) got tripped up.
It's kind of easy to for me to screw something up trying to pull bits
out of replies and such, but I think this one was pretty simple (looks
like maybe some PGP or MIME thing was the issue).

I just put
<https://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux.git/commit/?h=for-next&id=7699f7aacf3ebfee51c670b6f796b2797f0f7487>
on my staging branch, it looks OK to me but LMK if there's a problem.

Thanks!

>
> Best regards
> Uwe
>
> --
> Pengutronix e.K. | Uwe Kleine-König |
> Industrial Linux Solutions | https://www.pengutronix.de/ |

2022-06-01 07:18:06

by Uwe Kleine-König

[permalink] [raw]
Subject: Re: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

On Mon, May 30, 2022 at 04:11:27PM -0700, Palmer Dabbelt wrote:
> On Mon, 30 May 2022 12:41:33 PDT (-0700), [email protected] wrote:
> > Hello,
> >
> > On Mon, May 30, 2022 at 11:58:16AM -0700, Palmer Dabbelt wrote:
> > > On Mon, 30 May 2022 00:42:02 PDT (-0700), [email protected] wrote:
> > > > Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
> > > > commit 233c1e6c319c ("kexec_file: drop weak attribute from
> > > > arch_kexec_apply_relocations[_add]") is also contained in the tree.
> > > > This currently happens in next-20220527.
> > > > > Prepare the RISC-V similar to the s390 adaption done in
> > > 233c1e6c319c.
> > > > This is safe to do on top of the riscv change even without the change to
> > > > arch_kexec_apply_relocations.
> > > > > Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
> > > > Looks-good-to: liaochang (A) <[email protected]>
> > > > Signed-off-by: Uwe Kleine-K?nig <[email protected]>
> > > > ---
> > > > > On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
> > > > > > I can confirm that doing
> > > > > > > diff --git a/arch/riscv/include/asm/kexec.h
> > > > > b/arch/riscv/include/asm/kexec.h
> > > > > > index 206217b23301..eee260e8ab30 100644
> > > > > > --- a/arch/riscv/include/asm/kexec.h
> > > > > > +++ b/arch/riscv/include/asm/kexec.h
> > > > > > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
> > > > > > > #ifdef CONFIG_KEXEC_FILE
> > > > > > extern const struct kexec_file_ops elf_kexec_ops;
> > > > > > +
> > > > > > +struct purgatory_info;
> > > > > > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> > > > > > + Elf_Shdr *section,
> > > > > > + const Elf_Shdr *relsec,
> > > > > > + const Elf_Shdr *symtab);
> > > > > > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> > > > > > #endif
> > > > > > > #endif
> > > > > > > LGTM, you could send a fixup patch to riscv, thanks.
> > > > > > > > > on top of 838b3e28488f results in a compilable tree. And
> > > when
> > > > > merging
> > > > > > 233c1e6c319c into this, it is still building.
> > > > > > > I'm not enough into kexec (and riscv) to judge if this is
> > > > > sensible, or
> > > > > > create a useful commit log but the obvious way forward is to apply the
> > > > > > above patch to the riscv tree before it hits Linus' tree.
> > > > > Ok, here comes a patch with a generic commit log.
> > > > > @riscv people: If you prefer, squash it into 838b3e28488f.
> > >
> > > Sorry, just saw this after I sent my version of the fix. They're the same,
> > > but do you mind sending a full-on patch so I can merge it?
> >
> > Sorry, I don't understand your request. I found
> > https://lore.kernel.org/linux-riscv/[email protected]/
> >
> > but I don't know what a full-on patch is and what stops you merging my
> > patch.
> >
> > Is it that it's in reply to a patch series and b4 fails to fetch the
> > right patch and you ask to send it in a new thread?
>
> Ya, with the reply bits in there my mail merge (which unfortunately isn't
> b4, I haven't gotten around to converting yet) got tripped up. It's kind of
> easy to for me to screw something up trying to pull bits out of replies and
> such, but I think this one was pretty simple (looks like maybe some PGP or
> MIME thing was the issue).
>
> I just put <https://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux.git/commit/?h=for-next&id=7699f7aacf3ebfee51c670b6f796b2797f0f7487>
> on my staging branch, it looks OK to me but LMK if there's a problem.

there is indeed a problem, but caused by me: If you are ready to rewrite
it, please do s/week/weak/ in the subject line. Otherwise looks good to
me.

Best regards and thanks,
Uwe

--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (3.95 kB)
signature.asc (499.00 B)
Download all attachments

2022-06-01 19:57:14

by Uwe Kleine-König

[permalink] [raw]
Subject: Re: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

Hello,

On Mon, May 30, 2022 at 11:58:16AM -0700, Palmer Dabbelt wrote:
> On Mon, 30 May 2022 00:42:02 PDT (-0700), [email protected] wrote:
> > Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
> > commit 233c1e6c319c ("kexec_file: drop weak attribute from
> > arch_kexec_apply_relocations[_add]") is also contained in the tree.
> > This currently happens in next-20220527.
> >
> > Prepare the RISC-V similar to the s390 adaption done in 233c1e6c319c.
> > This is safe to do on top of the riscv change even without the change to
> > arch_kexec_apply_relocations.
> >
> > Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
> > Looks-good-to: liaochang (A) <[email protected]>
> > Signed-off-by: Uwe Kleine-K?nig <[email protected]>
> > ---
> >
> > On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
> > > > I can confirm that doing
> > > > > diff --git a/arch/riscv/include/asm/kexec.h
> > > b/arch/riscv/include/asm/kexec.h
> > > > index 206217b23301..eee260e8ab30 100644
> > > > --- a/arch/riscv/include/asm/kexec.h
> > > > +++ b/arch/riscv/include/asm/kexec.h
> > > > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
> > > > > #ifdef CONFIG_KEXEC_FILE
> > > > extern const struct kexec_file_ops elf_kexec_ops;
> > > > +
> > > > +struct purgatory_info;
> > > > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
> > > > + Elf_Shdr *section,
> > > > + const Elf_Shdr *relsec,
> > > > + const Elf_Shdr *symtab);
> > > > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
> > > > #endif
> > > > > #endif
> > >
> > > LGTM, you could send a fixup patch to riscv, thanks.
> > >
> > > > > on top of 838b3e28488f results in a compilable tree. And when
> > > merging
> > > > 233c1e6c319c into this, it is still building.
> > > > > I'm not enough into kexec (and riscv) to judge if this is
> > > sensible, or
> > > > create a useful commit log but the obvious way forward is to apply the
> > > > above patch to the riscv tree before it hits Linus' tree.
> >
> > Ok, here comes a patch with a generic commit log.
> >
> > @riscv people: If you prefer, squash it into 838b3e28488f.
>
> Sorry, just saw this after I sent my version of the fix. They're the same,
> but do you mind sending a full-on patch so I can merge it?

Sorry, I don't understand your request. I found
https://lore.kernel.org/linux-riscv/[email protected]/

but I don't know what a full-on patch is and what stops you merging my
patch.

Is it that it's in reply to a patch series and b4 fails to fetch the
right patch and you ask to send it in a new thread?

Best regards
Uwe

--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (2.89 kB)
signature.asc (499.00 B)
Download all attachments

2022-06-01 19:59:50

by Palmer Dabbelt

[permalink] [raw]
Subject: Re: [PATCH] RISC-V: Prepare dropping week attribute from arch_kexec_apply_relocations[_add]

On Mon, 30 May 2022 22:58:03 PDT (-0700), [email protected] wrote:
> On Mon, May 30, 2022 at 04:11:27PM -0700, Palmer Dabbelt wrote:
>> On Mon, 30 May 2022 12:41:33 PDT (-0700), [email protected] wrote:
>> > Hello,
>> >
>> > On Mon, May 30, 2022 at 11:58:16AM -0700, Palmer Dabbelt wrote:
>> > > On Mon, 30 May 2022 00:42:02 PDT (-0700), [email protected] wrote:
>> > > > Without this change arch/riscv/kernel/elf_kexec.c fails to compile once
>> > > > commit 233c1e6c319c ("kexec_file: drop weak attribute from
>> > > > arch_kexec_apply_relocations[_add]") is also contained in the tree.
>> > > > This currently happens in next-20220527.
>> > > > > Prepare the RISC-V similar to the s390 adaption done in
>> > > 233c1e6c319c.
>> > > > This is safe to do on top of the riscv change even without the change to
>> > > > arch_kexec_apply_relocations.
>> > > > > Fixes: 838b3e28488f ("RISC-V: Load purgatory in kexec_file")
>> > > > Looks-good-to: liaochang (A) <[email protected]>
>> > > > Signed-off-by: Uwe Kleine-König <[email protected]>
>> > > > ---
>> > > > > On Mon, May 30, 2022 at 09:43:26AM +0800, liaochang (A) wrote:
>> > > > > > I can confirm that doing
>> > > > > > > diff --git a/arch/riscv/include/asm/kexec.h
>> > > > > b/arch/riscv/include/asm/kexec.h
>> > > > > > index 206217b23301..eee260e8ab30 100644
>> > > > > > --- a/arch/riscv/include/asm/kexec.h
>> > > > > > +++ b/arch/riscv/include/asm/kexec.h
>> > > > > > @@ -55,6 +55,13 @@ extern riscv_kexec_method riscv_kexec_norelocate;
>> > > > > > > #ifdef CONFIG_KEXEC_FILE
>> > > > > > extern const struct kexec_file_ops elf_kexec_ops;
>> > > > > > +
>> > > > > > +struct purgatory_info;
>> > > > > > +int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
>> > > > > > + Elf_Shdr *section,
>> > > > > > + const Elf_Shdr *relsec,
>> > > > > > + const Elf_Shdr *symtab);
>> > > > > > +#define arch_kexec_apply_relocations_add arch_kexec_apply_relocations_add
>> > > > > > #endif
>> > > > > > > #endif
>> > > > > > > LGTM, you could send a fixup patch to riscv, thanks.
>> > > > > > > > > on top of 838b3e28488f results in a compilable tree. And
>> > > when
>> > > > > merging
>> > > > > > 233c1e6c319c into this, it is still building.
>> > > > > > > I'm not enough into kexec (and riscv) to judge if this is
>> > > > > sensible, or
>> > > > > > create a useful commit log but the obvious way forward is to apply the
>> > > > > > above patch to the riscv tree before it hits Linus' tree.
>> > > > > Ok, here comes a patch with a generic commit log.
>> > > > > @riscv people: If you prefer, squash it into 838b3e28488f.
>> > >
>> > > Sorry, just saw this after I sent my version of the fix. They're the same,
>> > > but do you mind sending a full-on patch so I can merge it?
>> >
>> > Sorry, I don't understand your request. I found
>> > https://lore.kernel.org/linux-riscv/[email protected]/
>> >
>> > but I don't know what a full-on patch is and what stops you merging my
>> > patch.
>> >
>> > Is it that it's in reply to a patch series and b4 fails to fetch the
>> > right patch and you ask to send it in a new thread?
>>
>> Ya, with the reply bits in there my mail merge (which unfortunately isn't
>> b4, I haven't gotten around to converting yet) got tripped up. It's kind of
>> easy to for me to screw something up trying to pull bits out of replies and
>> such, but I think this one was pretty simple (looks like maybe some PGP or
>> MIME thing was the issue).
>>
>> I just put <https://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux.git/commit/?h=for-next&id=7699f7aacf3ebfee51c670b6f796b2797f0f7487>
>> on my staging branch, it looks OK to me but LMK if there's a problem.
>
> there is indeed a problem, but caused by me: If you are ready to rewrite
> it, please do s/week/weak/ in the subject line. Otherwise looks good to
> me.

Ah, sorry, I missed that too. I was hoping to send a PR this morning,
I'd usually just rewrite and delay this one but given it's necessary to
make the post-merge build work I'm going to just send it up with the
spelling error.

>
> Best regards and thanks,
> Uwe
>
> --
> Pengutronix e.K. | Uwe Kleine-König |
> Industrial Linux Solutions | https://www.pengutronix.de/ |