[PATCH] arm64: mm: Fix dead assignment of old_pte

[Nathan Huckleberry]

When analyzed with the clang static analyzer the
following warning occurs

line 251, column 2
Value stored to 'old_pte' is never read

This warning is repeated every time pgtable.h is
included by another file and produces ~3500
extra warnings.

Moving old_pte into preprocessor guard.

Cc: clang-bu...@googlegroups.com
Signed-off-by: Nathan Huckleberry <nh...@google.com>
---
arch/arm64/include/asm/pgtable.h | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index fca26759081a..42ca4fc67f27 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -238,8 +238,6 @@ extern void __sync_icache_dcache(pte_t pteval);
static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
pte_t *ptep, pte_t pte)
{
- pte_t old_pte;
-
if (pte_present(pte) && pte_user_exec(pte) && !pte_special(pte))
__sync_icache_dcache(pte);

@@ -248,8 +246,11 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
* hardware updates of the pte (ptep_set_access_flags safely changes
* valid ptes without going through an invalid entry).
*/
+ #if IS_ENABLED(CONFIG_DEBUG_VM)
+ pte_t old_pte;
+
old_pte = READ_ONCE(*ptep);
- if (IS_ENABLED(CONFIG_DEBUG_VM) && pte_valid(old_pte) && pte_valid(pte) &&
+ if (pte_valid(old_pte) && pte_valid(pte) &&
(mm == current->active_mm || atomic_read(&mm->mm_users) > 1)) {
VM_WARN_ONCE(!pte_young(pte),
"%s: racy access flag clearing: 0x%016llx -> 0x%016llx",
@@ -258,6 +259,7 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
"%s: racy dirty state clearing: 0x%016llx -> 0x%016llx",
__func__, pte_val(old_pte), pte_val(pte));
}
+ #endif

set_pte(ptep, pte);
}
--
2.22.0.410.gd8fdbe21b5-goog

[Nathan Chancellor]

Hi Nathan,

This does not apply on -next because of https://git.kernel.org/arm64/c/9b604722059039a1a3ff69fb8dfd024264046024.
I would get into the habit of testing -next to see if the warning is
present there first because someone may have independently fixed it
already (I'd be surprised if it wasn't fixed by that commit from a quick
glance).

Additionally, when I do apply this patch to mainline and build, I see
the following warning:

In file included from /home/nathan/cbl/linux/arch/arm64/kernel/asm-offsets.c:10:
In file included from /home/nathan/cbl/linux/include/linux/arm_sdei.h:14:
In file included from /home/nathan/cbl/linux/include/acpi/ghes.h:5:
In file included from /home/nathan/cbl/linux/include/acpi/apei.h:9:
In file included from /home/nathan/cbl/linux/include/linux/acpi.h:34:
In file included from /home/nathan/cbl/linux/include/acpi/acpi_io.h:5:
In file included from /home/nathan/cbl/linux/include/linux/io.h:13:
In file included from /home/nathan/cbl/linux/arch/arm64/include/asm/io.h:18:
/home/nathan/cbl/linux/arch/arm64/include/asm/pgtable.h:250:8: warning: ISO C90 forbids mixing declarations and code [-Wdeclaration-after-statement]
pte_t old_pte;
^
1 warning generated.

Cheers,
Nathan

[Nathan Huckleberry]

Oops I forgot moving the variable declaration would cause a warning.
Will send a V2.

Thanks,
Nathan Huckleberry

> --
> You received this message because you are subscribed to the Google Groups "Clang Built Linux" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to clang-built-li...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/clang-built-linux/20190702232459.GA14941%40archlinux-epyc.

[Nathan Huckleberry]

When analyzed with the clang static analyzer the
following warning occurs

line 251, column 2
Value stored to 'old_pte' is never read

This warning is repeated every time pgtable.h is
included by another file and produces ~3500
extra warnings.

Moving old_pte into preprocessor guard.

Cc: clang-bu...@googlegroups.com
Signed-off-by: Nathan Huckleberry <nh...@google.com>
---

Changes from v1 -> v2
* Added scope to avoid [-Wdeclaration-after-statement]
arch/arm64/include/asm/pgtable.h | 27 ++++++++++++++++-----------
1 file changed, 16 insertions(+), 11 deletions(-)

diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index fca26759081a..12b7f08db40d 100644

--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -238,8 +238,6 @@ extern void __sync_icache_dcache(pte_t pteval);
static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
pte_t *ptep, pte_t pte)
{
- pte_t old_pte;
-
if (pte_present(pte) && pte_user_exec(pte) && !pte_special(pte))
__sync_icache_dcache(pte);

@@ -248,16 +246,23 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,

* hardware updates of the pte (ptep_set_access_flags safely changes
* valid ptes without going through an invalid entry).
*/

- old_pte = READ_ONCE(*ptep);

- if (IS_ENABLED(CONFIG_DEBUG_VM) && pte_valid(old_pte) && pte_valid(pte) &&

- (mm == current->active_mm || atomic_read(&mm->mm_users) > 1)) {
- VM_WARN_ONCE(!pte_young(pte),
- "%s: racy access flag clearing: 0x%016llx -> 0x%016llx",
- __func__, pte_val(old_pte), pte_val(pte));
- VM_WARN_ONCE(pte_write(old_pte) && !pte_dirty(pte),
- "%s: racy dirty state clearing: 0x%016llx -> 0x%016llx",
- __func__, pte_val(old_pte), pte_val(pte));
+ #if IS_ENABLED(CONFIG_DEBUG_VM)
+ {
+ pte_t old_pte;
+
+ old_pte = READ_ONCE(*ptep);

+ if (pte_valid(old_pte) && pte_valid(pte) &&

+ (mm == current->active_mm ||
+ atomic_read(&mm->mm_users) > 1)) {
+ VM_WARN_ONCE(!pte_young(pte),
+ "%s: racy access flag clearing: 0x%016llx -> 0x%016llx",
+ __func__, pte_val(old_pte), pte_val(pte));
+ VM_WARN_ONCE(pte_write(old_pte) && !pte_dirty(pte),
+ "%s: racy dirty state clearing: 0x%016llx -> 0x%016llx",
+ __func__, pte_val(old_pte), pte_val(pte));
+ }

[Vladimir Murzin]

On 7/3/19 12:41 AM, Nathan Huckleberry wrote:
> When analyzed with the clang static analyzer the
> following warning occurs
>
> line 251, column 2
> Value stored to 'old_pte' is never read
>
> This warning is repeated every time pgtable.h is
> included by another file and produces ~3500
> extra warnings.
>
> Moving old_pte into preprocessor guard.

I'm wondering if it is a case for __maybe_unused?

Something like:

- pte_t old_pte;
+ pte_t __maybe_unused old_pte;


Cheers
Vladimir

[Will Deacon]

On Tue, Jul 02, 2019 at 04:41:35PM -0700, Nathan Huckleberry wrote:
> When analyzed with the clang static analyzer the
> following warning occurs
>
> line 251, column 2
> Value stored to 'old_pte' is never read
>
> This warning is repeated every time pgtable.h is
> included by another file and produces ~3500
> extra warnings.

Does this warning actually trigger with linux-next?

Will

[Mark Rutland]

On Tue, Jul 02, 2019 at 04:41:35PM -0700, Nathan Huckleberry wrote:

> When analyzed with the clang static analyzer the
> following warning occurs
>
> line 251, column 2
> Value stored to 'old_pte' is never read
>
> This warning is repeated every time pgtable.h is
> included by another file and produces ~3500
> extra warnings.
>
> Moving old_pte into preprocessor guard.
>
> Cc: clang-bu...@googlegroups.com
> Signed-off-by: Nathan Huckleberry <nh...@google.com>
> ---
> Changes from v1 -> v2
> * Added scope to avoid [-Wdeclaration-after-statement]
> arch/arm64/include/asm/pgtable.h | 27 ++++++++++++++++-----------
> 1 file changed, 16 insertions(+), 11 deletions(-)

As Will asked, does this also trigger in linux-next?

I rewrote this code to avoid to only perform the READ_ONCE() when we'd
use the value, and IIUC that may be sufficient to avoid the warning:

https://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git/commit/?h=for-next/core&id=9b604722059039a1a3ff69fb8dfd024264046024

Thanks,
Mark.

> _______________________________________________
> linux-arm-kernel mailing list
> linux-ar...@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[Nathan Huckleberry]

Warning is not triggered in next. Looks like this patch is unnecessary.

Thanks,
Nathan Huckleberry