| Message ID | 1623860738-146761-1-git-send-email-quic_c_gdjako@quicinc.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mm/slub: Add taint after the errors are printed | expand |
On 6/16/21 6:25 PM, Georgi Djakov wrote: > When running the kernel with panic_on_taint, the usual slub debug error > messages are not being printed when object corruption happens. That's > because we panic in add_taint(), which is called before printing the > additional information. This is a bit unfortunate as the error messages > are actually very useful, especially before a panic. Let's fix this by > moving add_taint() after the errors are printed on the console. > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> Makes sense. While at it, I wonder if we should use LOCKDEP_STILL_OK instead of LOCKDEP_NOW_UNRELIABLE. Isn't it too pessimistic to assume that some slab's memory corruption hit some lock state? > --- > mm/slub.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/mm/slub.c b/mm/slub.c > index a8b0daa1a307..ce7b8e4551b5 100644 > --- a/mm/slub.c > +++ b/mm/slub.c > @@ -719,8 +719,6 @@ static void slab_bug(struct kmem_cache *s, char *fmt, ...) > pr_err("=============================================================================\n"); > pr_err("BUG %s (%s): %pV\n", s->name, print_tainted(), &vaf); > pr_err("-----------------------------------------------------------------------------\n\n"); > - > - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > va_end(args); > } > > @@ -801,6 +799,7 @@ void object_err(struct kmem_cache *s, struct page *page, > > slab_bug(s, "%s", reason); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > @@ -818,6 +817,7 @@ static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > slab_bug(s, "%s", buf); > print_page_info(page); > dump_stack(); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static void init_object(struct kmem_cache *s, void *object, u8 val) > @@ -869,6 +869,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page, > fault, end - 1, fault - addr, > fault[0], value); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > > skip_bug_print: > restore_bytes(s, what, value, fault, end); >
On Wed, Jun 16, 2021 at 09:25:38AM -0700, Georgi Djakov wrote: > When running the kernel with panic_on_taint, the usual slub debug error > messages are not being printed when object corruption happens. That's > because we panic in add_taint(), which is called before printing the > additional information. This is a bit unfortunate as the error messages > are actually very useful, especially before a panic. Let's fix this by > moving add_taint() after the errors are printed on the console. > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> > --- > mm/slub.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/mm/slub.c b/mm/slub.c > index a8b0daa1a307..ce7b8e4551b5 100644 > --- a/mm/slub.c > +++ b/mm/slub.c > @@ -719,8 +719,6 @@ static void slab_bug(struct kmem_cache *s, char *fmt, ...) > pr_err("=============================================================================\n"); > pr_err("BUG %s (%s): %pV\n", s->name, print_tainted(), &vaf); > pr_err("-----------------------------------------------------------------------------\n\n"); > - > - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > va_end(args); > } > > @@ -801,6 +799,7 @@ void object_err(struct kmem_cache *s, struct page *page, > > slab_bug(s, "%s", reason); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > @@ -818,6 +817,7 @@ static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > slab_bug(s, "%s", buf); > print_page_info(page); > dump_stack(); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static void init_object(struct kmem_cache *s, void *object, u8 val) > @@ -869,6 +869,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page, > fault, end - 1, fault - addr, > fault[0], value); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > > skip_bug_print: > restore_bytes(s, what, value, fault, end); > Acked-by: Rafael Aquini <aquini@redhat.com>
On Wed, Jun 16, 2021 at 06:34:41PM +0200, Vlastimil Babka wrote: > On 6/16/21 6:25 PM, Georgi Djakov wrote: > > When running the kernel with panic_on_taint, the usual slub debug error > > messages are not being printed when object corruption happens. That's > > because we panic in add_taint(), which is called before printing the > > additional information. This is a bit unfortunate as the error messages > > are actually very useful, especially before a panic. Let's fix this by > > moving add_taint() after the errors are printed on the console. > > > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> > > Makes sense. > > While at it, I wonder if we should use LOCKDEP_STILL_OK instead of > LOCKDEP_NOW_UNRELIABLE. Isn't it too pessimistic to assume that some slab's > memory corruption hit some lock state? > Given there is noted corruption I don't think it's safe to assume otherwise.
On Wed 2021-06-16 09:25 -0700, Georgi Djakov wrote: > When running the kernel with panic_on_taint, the usual slub debug error > messages are not being printed when object corruption happens. That's > because we panic in add_taint(), which is called before printing the > additional information. This is a bit unfortunate as the error messages > are actually very useful, especially before a panic. Let's fix this by > moving add_taint() after the errors are printed on the console. > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> > --- > mm/slub.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/mm/slub.c b/mm/slub.c > index a8b0daa1a307..ce7b8e4551b5 100644 > --- a/mm/slub.c > +++ b/mm/slub.c > @@ -719,8 +719,6 @@ static void slab_bug(struct kmem_cache *s, char *fmt, ...) > pr_err("=============================================================================\n"); > pr_err("BUG %s (%s): %pV\n", s->name, print_tainted(), &vaf); > pr_err("-----------------------------------------------------------------------------\n\n"); > - > - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > va_end(args); > } > > @@ -801,6 +799,7 @@ void object_err(struct kmem_cache *s, struct page *page, > > slab_bug(s, "%s", reason); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > @@ -818,6 +817,7 @@ static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > slab_bug(s, "%s", buf); > print_page_info(page); > dump_stack(); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static void init_object(struct kmem_cache *s, void *object, u8 val) > @@ -869,6 +869,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page, > fault, end - 1, fault - addr, > fault[0], value); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > > skip_bug_print: > restore_bytes(s, what, value, fault, end); > Good catch. Thanks! Reviewed-by: Aaron Tomlin <atomlin@redhat.com>
On 6/16/21 6:25 PM, Georgi Djakov wrote: > When running the kernel with panic_on_taint, the usual slub debug error > messages are not being printed when object corruption happens. That's > because we panic in add_taint(), which is called before printing the > additional information. This is a bit unfortunate as the error messages > are actually very useful, especially before a panic. Let's fix this by > moving add_taint() after the errors are printed on the console. > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> Acked-by: Vlastimil Babka <vbabka@suse.cz> > --- > mm/slub.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/mm/slub.c b/mm/slub.c > index a8b0daa1a307..ce7b8e4551b5 100644 > --- a/mm/slub.c > +++ b/mm/slub.c > @@ -719,8 +719,6 @@ static void slab_bug(struct kmem_cache *s, char *fmt, ...) > pr_err("=============================================================================\n"); > pr_err("BUG %s (%s): %pV\n", s->name, print_tainted(), &vaf); > pr_err("-----------------------------------------------------------------------------\n\n"); > - > - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > va_end(args); > } > > @@ -801,6 +799,7 @@ void object_err(struct kmem_cache *s, struct page *page, > > slab_bug(s, "%s", reason); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > @@ -818,6 +817,7 @@ static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, > slab_bug(s, "%s", buf); > print_page_info(page); > dump_stack(); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > } > > static void init_object(struct kmem_cache *s, void *object, u8 val) > @@ -869,6 +869,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page, > fault, end - 1, fault - addr, > fault[0], value); > print_trailer(s, page, object); > + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); > > skip_bug_print: > restore_bytes(s, what, value, fault, end); >
On Wed, 16 Jun 2021, Georgi Djakov wrote: > When running the kernel with panic_on_taint, the usual slub debug error > messages are not being printed when object corruption happens. That's > because we panic in add_taint(), which is called before printing the > additional information. This is a bit unfortunate as the error messages > are actually very useful, especially before a panic. Let's fix this by > moving add_taint() after the errors are printed on the console. > > Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> Acked-by: David Rientjes <rientjes@google.com>
diff --git a/mm/slub.c b/mm/slub.c index a8b0daa1a307..ce7b8e4551b5 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -719,8 +719,6 @@ static void slab_bug(struct kmem_cache *s, char *fmt, ...) pr_err("=============================================================================\n"); pr_err("BUG %s (%s): %pV\n", s->name, print_tainted(), &vaf); pr_err("-----------------------------------------------------------------------------\n\n"); - - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); va_end(args); } @@ -801,6 +799,7 @@ void object_err(struct kmem_cache *s, struct page *page, slab_bug(s, "%s", reason); print_trailer(s, page, object); + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); } static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, @@ -818,6 +817,7 @@ static __printf(3, 4) void slab_err(struct kmem_cache *s, struct page *page, slab_bug(s, "%s", buf); print_page_info(page); dump_stack(); + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); } static void init_object(struct kmem_cache *s, void *object, u8 val) @@ -869,6 +869,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page, fault, end - 1, fault - addr, fault[0], value); print_trailer(s, page, object); + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); skip_bug_print: restore_bytes(s, what, value, fault, end);
When running the kernel with panic_on_taint, the usual slub debug error messages are not being printed when object corruption happens. That's because we panic in add_taint(), which is called before printing the additional information. This is a bit unfortunate as the error messages are actually very useful, especially before a panic. Let's fix this by moving add_taint() after the errors are printed on the console. Signed-off-by: Georgi Djakov <quic_c_gdjako@quicinc.com> --- mm/slub.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)