| Message ID | 20190702234135.78780-1-nhuck@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] arm64: mm: Fix dead assignment of old_pte | expand |
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 > > Cc: clang-built-linux@googlegroups.com > Signed-off-by: Nathan Huckleberry <nhuck@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)); > + } > } > + #endif > > set_pte(ptep, pte); > } >
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
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-built-linux@googlegroups.com > Signed-off-by: Nathan Huckleberry <nhuck@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. > > 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)); > + } > } > + #endif > > set_pte(ptep, pte); > } > -- > 2.22.0.410.gd8fdbe21b5-goog > > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Warning is not triggered in next. Looks like this patch is unnecessary. Thanks, Nathan Huckleberry On Wed, Jul 3, 2019 at 4:23 AM Mark Rutland <mark.rutland@arm.com> wrote: > > 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-built-linux@googlegroups.com > > Signed-off-by: Nathan Huckleberry <nhuck@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. > > > > > 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)); > > + } > > } > > + #endif > > > > set_pte(ptep, pte); > > } > > -- > > 2.22.0.410.gd8fdbe21b5-goog > > > > > > _______________________________________________ > > linux-arm-kernel mailing list > > linux-arm-kernel@lists.infradead.org > > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
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)); + } } + #endif set_pte(ptep, pte); }
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-built-linux@googlegroups.com Signed-off-by: Nathan Huckleberry <nhuck@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(-)