| Message ID | 20150211144319.2a0c29d6@cuia.bos.redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Feb 11, 2015 at 02:43:19PM -0500, Rik van Riel wrote: > If exception_enter happens when already in IN_KERNEL state, the > code still calls context_tracking_exit, which ends up in > rcu_eqs_exit_common, which explodes with a WARN_ON when it is > called in a situation where dynticks are not enabled. > > This can be avoided by having exception_enter only switch to > IN_KERNEL state if the current state is not already IN_KERNEL. Ugh... Time to formally verify, sounds like... Thanx, Paul > Signed-off-by: Rik van Riel <riel@redhat.com> > Reported-by: Luiz Capitulino <lcapitulino@redhat.com> > --- > Frederic, you will want this "bonus" patch, too :) > > Thanks to Luiz for finding this one. Whatever I was running did not > trigger this issue... > > include/linux/context_tracking.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/include/linux/context_tracking.h b/include/linux/context_tracking.h > index b65fd1420e53..9da230406e8c 100644 > --- a/include/linux/context_tracking.h > +++ b/include/linux/context_tracking.h > @@ -37,7 +37,8 @@ static inline enum ctx_state exception_enter(void) > return 0; > > prev_ctx = this_cpu_read(context_tracking.state); > - context_tracking_exit(prev_ctx); > + if (prev_ctx != IN_KERNEL) > + context_tracking_exit(prev_ctx); > > return prev_ctx; > } > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Feb 11, 2015 at 02:43:19PM -0500, Rik van Riel wrote: > If exception_enter happens when already in IN_KERNEL state, the > code still calls context_tracking_exit, which ends up in > rcu_eqs_exit_common, which explodes with a WARN_ON when it is > called in a situation where dynticks are not enabled. Fortunately context_tracking_exit() already has a current_state == IN_KERNEL check so this shouldn't be a problem. Meanwhile I'll still take the patch, it's better to handle that from the caller. Thanks. > > This can be avoided by having exception_enter only switch to > IN_KERNEL state if the current state is not already IN_KERNEL. > > Signed-off-by: Rik van Riel <riel@redhat.com> > Reported-by: Luiz Capitulino <lcapitulino@redhat.com> > --- > Frederic, you will want this "bonus" patch, too :) > > Thanks to Luiz for finding this one. Whatever I was running did not > trigger this issue... > > include/linux/context_tracking.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/include/linux/context_tracking.h b/include/linux/context_tracking.h > index b65fd1420e53..9da230406e8c 100644 > --- a/include/linux/context_tracking.h > +++ b/include/linux/context_tracking.h > @@ -37,7 +37,8 @@ static inline enum ctx_state exception_enter(void) > return 0; > > prev_ctx = this_cpu_read(context_tracking.state); > - context_tracking_exit(prev_ctx); > + if (prev_ctx != IN_KERNEL) > + context_tracking_exit(prev_ctx); > > return prev_ctx; > } > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 02/12/2015 10:42 AM, Frederic Weisbecker wrote: > On Wed, Feb 11, 2015 at 02:43:19PM -0500, Rik van Riel wrote: >> If exception_enter happens when already in IN_KERNEL state, the >> code still calls context_tracking_exit, which ends up in >> rcu_eqs_exit_common, which explodes with a WARN_ON when it is >> called in a situation where dynticks are not enabled. > > Fortunately context_tracking_exit() already has a current_state == > IN_KERNEL check so this shouldn't be a problem. No, it had a hard-coded "current_state == IN_USER" check, which is very close, but ... ... I replaced that with a state argument, and forgot to ensure that it never gets called with state == IN_KERNEL. This patch fixes that. > Meanwhile I'll still take the patch, it's better to handle that > from the caller. Thanks. - -- All rights reversed -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJU3Mr+AAoJEM553pKExN6DYNUH/2m9CtXhLdTHOEHRvxg41PCZ /xafetUOS9cka0CNuiYpUuvfMSucoePW7YqUXqjYSIP25DsAleOh0qdep1Ob5bH+ 2BqZNMwK3QDHf1+/V7nulnjVkeHtpXJm0HIZOjc06xeL+9T6ydB1vhQGIMLrGL9S LvOstI3fseeIgglwYc2Gx7H7e99oOkxysvwMMvcMrW0cPSRAOdYxINQnfYW8A5kq DTTXwWuJRZa4FLtP3wLpvocm5dMGDwTsDmuOk1PmXYlsTsO6H2BmCeio0euzStoJ l+jR4x7Aq2KXES7gnMgpPw1iON3xKJ/RbXF8IC/doII8FYEV8Raxnf7hl47etBw= =yIjW -----END PGP SIGNATURE----- -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Feb 12, 2015 at 10:47:10AM -0500, Rik van Riel wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On 02/12/2015 10:42 AM, Frederic Weisbecker wrote: > > On Wed, Feb 11, 2015 at 02:43:19PM -0500, Rik van Riel wrote: > >> If exception_enter happens when already in IN_KERNEL state, the > >> code still calls context_tracking_exit, which ends up in > >> rcu_eqs_exit_common, which explodes with a WARN_ON when it is > >> called in a situation where dynticks are not enabled. > > > > Fortunately context_tracking_exit() already has a current_state == > > IN_KERNEL check so this shouldn't be a problem. > > No, it had a hard-coded "current_state == IN_USER" check, > which is very close, but ... > > ... I replaced that with a state argument, and forgot to > ensure that it never gets called with state == IN_KERNEL. > This patch fixes that. Ah that's right! Well I'm going to merge this patch to 1/5 then to avoid breaking bisection. Thanks. -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 02/12/2015 12:00 PM, Frederic Weisbecker wrote: > On Thu, Feb 12, 2015 at 10:47:10AM -0500, Rik van Riel wrote: >> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 >> >> On 02/12/2015 10:42 AM, Frederic Weisbecker wrote: >>> On Wed, Feb 11, 2015 at 02:43:19PM -0500, Rik van Riel wrote: >>>> If exception_enter happens when already in IN_KERNEL state, >>>> the code still calls context_tracking_exit, which ends up in >>>> rcu_eqs_exit_common, which explodes with a WARN_ON when it >>>> is called in a situation where dynticks are not enabled. >>> >>> Fortunately context_tracking_exit() already has a current_state >>> == IN_KERNEL check so this shouldn't be a problem. >> >> No, it had a hard-coded "current_state == IN_USER" check, which >> is very close, but ... >> >> ... I replaced that with a state argument, and forgot to ensure >> that it never gets called with state == IN_KERNEL. This patch >> fixes that. > > Ah that's right! Well I'm going to merge this patch to 1/5 then to > avoid breaking bisection. Thank you, Frederic! - -- All rights reversed -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJU3OeDAAoJEM553pKExN6D7BsIAJ8CKC73jQ8T5Dqa/tlHV7Db QFSJdpxP+7jCZwssehgpjpxCwtJ0UvGgle5OwX/POUhmagHxHmxVydOBz+xfYdBr UuGkEl5TL+oyoMUr80Q4RTnJSZN08zi+THqiv33tyPUj6cNiycBZAuho3ELTRNOA bRcHrMW+xd95uqoung7dSKrgA2jcym3+umNGnQb0gniraqcNLAmWs+jfAO8yZLJg vk8bIKed6epQ3n6gcdYe0A28cLOuBvjEs5JNcEPxujY/349sjitKR2pLQ6HsfHLV frlKsh7qQIRtoUJLO9ZBBDtGrmThwBwH8rw+GcVR8zviPNvV4IRrx47VBcHDWjc= =mwFO -----END PGP SIGNATURE----- -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/include/linux/context_tracking.h b/include/linux/context_tracking.h index b65fd1420e53..9da230406e8c 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -37,7 +37,8 @@ static inline enum ctx_state exception_enter(void) return 0; prev_ctx = this_cpu_read(context_tracking.state); - context_tracking_exit(prev_ctx); + if (prev_ctx != IN_KERNEL) + context_tracking_exit(prev_ctx); return prev_ctx; }
If exception_enter happens when already in IN_KERNEL state, the code still calls context_tracking_exit, which ends up in rcu_eqs_exit_common, which explodes with a WARN_ON when it is called in a situation where dynticks are not enabled. This can be avoided by having exception_enter only switch to IN_KERNEL state if the current state is not already IN_KERNEL. Signed-off-by: Rik van Riel <riel@redhat.com> Reported-by: Luiz Capitulino <lcapitulino@redhat.com> --- Frederic, you will want this "bonus" patch, too :) Thanks to Luiz for finding this one. Whatever I was running did not trigger this issue... include/linux/context_tracking.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html