| Message ID | 20241122071630.63707-3-guwen@linux.alibaba.com (mailing list archive) |
|---|---|
| State | New |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | two fixes for SMC | expand |
On 22.11.24 08:16, Wen Gu wrote: > We encountered a LGR/link use-after-free issue, which manifested as > the LGR/link refcnt reaching 0 early and entering the clear process, > making resource access unsafe. > How did you make sure that the refcount mentioned in the warning are the LGR/link refcnt, not &sk->sk_refcnt? > refcount_t: addition on 0; use-after-free. > WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140 > Workqueue: events smc_lgr_terminate_work [smc] > Call trace: > refcount_warn_saturate+0x9c/0x140 > __smc_lgr_terminate.part.45+0x2a8/0x370 [smc] > smc_lgr_terminate_work+0x28/0x30 [smc] > process_one_work+0x1b8/0x420 > worker_thread+0x158/0x510 > kthread+0x114/0x118 > > or > > refcount_t: underflow; use-after-free. > WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140 > Workqueue: smc_hs_wq smc_listen_work [smc] > Call trace: > refcount_warn_saturate+0xf0/0x140 > smcr_link_put+0x1cc/0x1d8 [smc] > smc_conn_free+0x110/0x1b0 [smc] > smc_conn_abort+0x50/0x60 [smc] > smc_listen_find_device+0x75c/0x790 [smc] > smc_listen_work+0x368/0x8a0 [smc] > process_one_work+0x1b8/0x420 > worker_thread+0x158/0x510 > kthread+0x114/0x118 > > It is caused by repeated release of LGR/link refcnt. One suspect is that > smc_conn_free() is called repeatedly because some smc_conn_free() are not > protected by sock lock. > > Calls under socklock | Calls not under socklock > ------------------------------------------------------- > lock_sock(sk) | smc_conn_abort > smc_conn_free | \- smc_conn_free > \- smcr_link_put | \- smcr_link_put (duplicated) > release_sock(sk) > > So make sure smc_conn_free() is called under the sock lock. > If I understand correctly, the fix could only solve a part of the problem, i.e. what the second call trace reported, right? > Fixes: 8cf3f3e42374 ("net/smc: use helper smc_conn_abort() in listen processing") > Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> > Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> > Co-developed-by: Kai <KaiShen@linux.alibaba.com> > Signed-off-by: Kai <KaiShen@linux.alibaba.com> > Signed-off-by: Wen Gu <guwen@linux.alibaba.com> > --- > net/smc/af_smc.c | 25 +++++++++++++++++++++---- > 1 file changed, 21 insertions(+), 4 deletions(-) > > diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c > index ed6d4d520bc7..e0a7a0151b11 100644 > --- a/net/smc/af_smc.c > +++ b/net/smc/af_smc.c > @@ -973,7 +973,8 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, > return smc_connect_fallback(smc, reason_code); > } > > -static void smc_conn_abort(struct smc_sock *smc, int local_first) > +static void __smc_conn_abort(struct smc_sock *smc, int local_first, > + bool locked) > { > struct smc_connection *conn = &smc->conn; > struct smc_link_group *lgr = conn->lgr; > @@ -982,11 +983,27 @@ static void smc_conn_abort(struct smc_sock *smc, int local_first) > if (smc_conn_lgr_valid(conn)) > lgr_valid = true; > > - smc_conn_free(conn); > + if (!locked) { > + lock_sock(&smc->sk); > + smc_conn_free(conn); > + release_sock(&smc->sk); > + } else { > + smc_conn_free(conn); > + } > if (local_first && lgr_valid) > smc_lgr_cleanup_early(lgr); > } > > +static void smc_conn_abort(struct smc_sock *smc, int local_first) > +{ > + __smc_conn_abort(smc, local_first, false); > +} > + > +static void smc_conn_abort_locked(struct smc_sock *smc, int local_first) > +{ > + __smc_conn_abort(smc, local_first, true); > +} > + > /* check if there is a rdma device available for this connection. */ > /* called for connect and listen */ > static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) > @@ -1352,7 +1369,7 @@ static int smc_connect_rdma(struct smc_sock *smc, > > return 0; > connect_abort: > - smc_conn_abort(smc, ini->first_contact_local); > + smc_conn_abort_locked(smc, ini->first_contact_local); > mutex_unlock(&smc_client_lgr_pending); > smc->connect_nonblock = 0; > > @@ -1454,7 +1471,7 @@ static int smc_connect_ism(struct smc_sock *smc, > > return 0; > connect_abort: > - smc_conn_abort(smc, ini->first_contact_local); > + smc_conn_abort_locked(smc, ini->first_contact_local); > mutex_unlock(&smc_server_lgr_pending); > smc->connect_nonblock = 0; > Why is smc_conn_abort_locked() only necessary for the smc_connect_work, not for the smc_listen_work? Thanks, Wenjia
On 22.11.24 08:16, Wen Gu wrote: > We encountered a LGR/link use-after-free issue, which manifested as > the LGR/link refcnt reaching 0 early and entering the clear process, > making resource access unsafe. > > refcount_t: addition on 0; use-after-free. > WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140 > Workqueue: events smc_lgr_terminate_work [smc] > Call trace: > refcount_warn_saturate+0x9c/0x140 > __smc_lgr_terminate.part.45+0x2a8/0x370 [smc] > smc_lgr_terminate_work+0x28/0x30 [smc] > process_one_work+0x1b8/0x420 > worker_thread+0x158/0x510 > kthread+0x114/0x118 > > or > > refcount_t: underflow; use-after-free. > WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140 > Workqueue: smc_hs_wq smc_listen_work [smc] > Call trace: > refcount_warn_saturate+0xf0/0x140 > smcr_link_put+0x1cc/0x1d8 [smc] > smc_conn_free+0x110/0x1b0 [smc] > smc_conn_abort+0x50/0x60 [smc] > smc_listen_find_device+0x75c/0x790 [smc] > smc_listen_work+0x368/0x8a0 [smc] > process_one_work+0x1b8/0x420 > worker_thread+0x158/0x510 > kthread+0x114/0x118 > > It is caused by repeated release of LGR/link refcnt. One suspect is that > smc_conn_free() is called repeatedly because some smc_conn_free() are not > protected by sock lock. > > Calls under socklock | Calls not under socklock > ------------------------------------------------------- > lock_sock(sk) | smc_conn_abort > smc_conn_free | \- smc_conn_free > \- smcr_link_put | \- smcr_link_put (duplicated) > release_sock(sk) > > So make sure smc_conn_free() is called under the sock lock. > > Fixes: 8cf3f3e42374 ("net/smc: use helper smc_conn_abort() in listen processing") > Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> > Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> > Co-developed-by: Kai <KaiShen@linux.alibaba.com> > Signed-off-by: Kai <KaiShen@linux.alibaba.com> > Signed-off-by: Wen Gu <guwen@linux.alibaba.com> > --- > net/smc/af_smc.c | 25 +++++++++++++++++++++---- > 1 file changed, 21 insertions(+), 4 deletions(-) > > diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c > index ed6d4d520bc7..e0a7a0151b11 100644 > --- a/net/smc/af_smc.c > +++ b/net/smc/af_smc.c > @@ -973,7 +973,8 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, > return smc_connect_fallback(smc, reason_code); > } > > -static void smc_conn_abort(struct smc_sock *smc, int local_first) > +static void __smc_conn_abort(struct smc_sock *smc, int local_first, > + bool locked) > { > struct smc_connection *conn = &smc->conn; > struct smc_link_group *lgr = conn->lgr; > @@ -982,11 +983,27 @@ static void smc_conn_abort(struct smc_sock *smc, int local_first) > if (smc_conn_lgr_valid(conn)) > lgr_valid = true; > > - smc_conn_free(conn); > + if (!locked) { > + lock_sock(&smc->sk); > + smc_conn_free(conn); > + release_sock(&smc->sk); > + } else { > + smc_conn_free(conn); > + } > if (local_first && lgr_valid) > smc_lgr_cleanup_early(lgr); > } > > +static void smc_conn_abort(struct smc_sock *smc, int local_first) > +{ > + __smc_conn_abort(smc, local_first, false); > +} > + > +static void smc_conn_abort_locked(struct smc_sock *smc, int local_first) > +{ > + __smc_conn_abort(smc, local_first, true); > +} > + > /* check if there is a rdma device available for this connection. */ > /* called for connect and listen */ > static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) > @@ -1352,7 +1369,7 @@ static int smc_connect_rdma(struct smc_sock *smc, > > return 0; > connect_abort: > - smc_conn_abort(smc, ini->first_contact_local); > + smc_conn_abort_locked(smc, ini->first_contact_local); > mutex_unlock(&smc_client_lgr_pending); > smc->connect_nonblock = 0; > > @@ -1454,7 +1471,7 @@ static int smc_connect_ism(struct smc_sock *smc, > > return 0; > connect_abort: > - smc_conn_abort(smc, ini->first_contact_local); > + smc_conn_abort_locked(smc, ini->first_contact_local); > mutex_unlock(&smc_server_lgr_pending); > smc->connect_nonblock = 0; > I wonder if this can deadlock, when you take lock_sock so far down in the callchain. example: smc_connect will first take lock_sock(sk) and then mutex_lock(&smc_server_lgr_pending); (e.g. in smc_connect_ism()) wheras smc_listen_work() will take mutex_lock(&smc_server_lgr_pending); and then lock_sock(sk) (in your __smc_conn_abort(,,false)) I am not sure whether this can be called on the same socket, but it looks suspicious to me. All callers of smc_conn_abort() without socklock seem to originate from smc_listen_work(). That makes me think whether smc_listen_work() should do lock_sock(sk) on a higher level. Do you have an example which function could collide with smc_listen_work()? i.e. have you found a way to reproduce this? Are you sure that all callers of smc_conn_free(), that are not smc_conn_abort(), do set the socklock? It seems to me that the path of smc_conn_kill() is not covered by your solution. Please excuse, that I am not deeply familiar with this code. I'm just trying to ask helpful questions.
On 22.11.24 17:03, Alexandra Winter wrote: > Are you sure that all callers of smc_conn_free(), that are not smc_conn_abort(), do set the socklock? > It seems to me that the path of smc_conn_kill() is not covered by your solution. My bad. smc_conn_kill() is called under socklock
On 2024/11/22 23:56, Wenjia Zhang wrote: > > > On 22.11.24 08:16, Wen Gu wrote: >> We encountered a LGR/link use-after-free issue, which manifested as >> the LGR/link refcnt reaching 0 early and entering the clear process, >> making resource access unsafe. >> > > How did you make sure that the refcount mentioned in the warning are the LGR/link refcnt, not &sk->sk_refcnt? > Because according to the panic stack, the UAF is found in smcr_link_put(), and it is also found that the link has been cleared at that time (lnk has been memset to all zero by __smcr_link_clear()). >> refcount_t: addition on 0; use-after-free. >> WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140 >> Workqueue: events smc_lgr_terminate_work [smc] >> Call trace: >> refcount_warn_saturate+0x9c/0x140 >> __smc_lgr_terminate.part.45+0x2a8/0x370 [smc] >> smc_lgr_terminate_work+0x28/0x30 [smc] >> process_one_work+0x1b8/0x420 >> worker_thread+0x158/0x510 >> kthread+0x114/0x118 >> >> or >> >> refcount_t: underflow; use-after-free. >> WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140 >> Workqueue: smc_hs_wq smc_listen_work [smc] >> Call trace: >> refcount_warn_saturate+0xf0/0x140 >> smcr_link_put+0x1cc/0x1d8 [smc] >> smc_conn_free+0x110/0x1b0 [smc] >> smc_conn_abort+0x50/0x60 [smc] >> smc_listen_find_device+0x75c/0x790 [smc] >> smc_listen_work+0x368/0x8a0 [smc] >> process_one_work+0x1b8/0x420 >> worker_thread+0x158/0x510 >> kthread+0x114/0x118 >> >> It is caused by repeated release of LGR/link refcnt. One suspect is that >> smc_conn_free() is called repeatedly because some smc_conn_free() are not >> protected by sock lock. >> >> Calls under socklock | Calls not under socklock >> ------------------------------------------------------- >> lock_sock(sk) | smc_conn_abort >> smc_conn_free | \- smc_conn_free >> \- smcr_link_put | \- smcr_link_put (duplicated) >> release_sock(sk) >> >> So make sure smc_conn_free() is called under the sock lock. >> > > If I understand correctly, the fix could only solve a part of the problem, i.e. what the second call trace reported, right? I think that these panic stacks (there are some other variations that I haven't posted) have the same root cause, that is the link/lgr refcnt reaches 0 early in the race situation, making access to link/lgr related resources no longer safe. The link/lgr refcnt was introduced by [1] & [2], the link refcnt is operated by link itself and connections registered to it, and the lgr refcnt is operated by lgr itself, links belong to it and connections registered to it. Through code analysis, the most likely suspect is that smc_conn_free() duplicate put link/lgr refcnt because some smc_conn_free() calls by smc_conn_abort() are not under the protection of sock lock, so if they are called at the same time, there may be a race condition. for example: __smc_lgr_terminate | smc_listen_decline -------------------------------------------------------------- lock_sock | smc_conn_kill | smc_conn_abort \- smc_conn_free | \- smc_conn_free release_sock | [1] 61f434b0280e ("net/smc: Resolve the race between link group access and termination") [2] 20c9398d3309 ("net/smc: Resolve the race between SMC-R link access and clear") > >> Fixes: 8cf3f3e42374 ("net/smc: use helper smc_conn_abort() in listen processing") >> Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> >> Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> >> Co-developed-by: Kai <KaiShen@linux.alibaba.com> >> Signed-off-by: Kai <KaiShen@linux.alibaba.com> >> Signed-off-by: Wen Gu <guwen@linux.alibaba.com> >> --- >> net/smc/af_smc.c | 25 +++++++++++++++++++++---- >> 1 file changed, 21 insertions(+), 4 deletions(-) >> >> diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c >> index ed6d4d520bc7..e0a7a0151b11 100644 >> --- a/net/smc/af_smc.c >> +++ b/net/smc/af_smc.c >> @@ -973,7 +973,8 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, >> return smc_connect_fallback(smc, reason_code); >> } >> -static void smc_conn_abort(struct smc_sock *smc, int local_first) >> +static void __smc_conn_abort(struct smc_sock *smc, int local_first, >> + bool locked) >> { >> struct smc_connection *conn = &smc->conn; >> struct smc_link_group *lgr = conn->lgr; >> @@ -982,11 +983,27 @@ static void smc_conn_abort(struct smc_sock *smc, int local_first) >> if (smc_conn_lgr_valid(conn)) >> lgr_valid = true; >> - smc_conn_free(conn); >> + if (!locked) { >> + lock_sock(&smc->sk); >> + smc_conn_free(conn); >> + release_sock(&smc->sk); >> + } else { >> + smc_conn_free(conn); >> + } >> if (local_first && lgr_valid) >> smc_lgr_cleanup_early(lgr); >> } >> +static void smc_conn_abort(struct smc_sock *smc, int local_first) >> +{ >> + __smc_conn_abort(smc, local_first, false); >> +} >> + >> +static void smc_conn_abort_locked(struct smc_sock *smc, int local_first) >> +{ >> + __smc_conn_abort(smc, local_first, true); >> +} >> + >> /* check if there is a rdma device available for this connection. */ >> /* called for connect and listen */ >> static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) >> @@ -1352,7 +1369,7 @@ static int smc_connect_rdma(struct smc_sock *smc, >> return 0; >> connect_abort: >> - smc_conn_abort(smc, ini->first_contact_local); >> + smc_conn_abort_locked(smc, ini->first_contact_local); >> mutex_unlock(&smc_client_lgr_pending); >> smc->connect_nonblock = 0; >> @@ -1454,7 +1471,7 @@ static int smc_connect_ism(struct smc_sock *smc, >> return 0; >> connect_abort: >> - smc_conn_abort(smc, ini->first_contact_local); >> + smc_conn_abort_locked(smc, ini->first_contact_local); >> mutex_unlock(&smc_server_lgr_pending); >> smc->connect_nonblock = 0; > > Why is smc_conn_abort_locked() only necessary for the smc_connect_work, not for the smc_listen_work? > Before this patch, the smc_conn_abort()->smc_conn_free() calls are not protected by sock lock except in smc_conn_{rdma|ism}. So I add sock lock protection inside the __smc_conn_abort() and introduce smc_conn_abort_locked() (which means sock lock has been taken) for smc_conn_{rdma|ism}. > Thanks, > Wenjia Thanks! Wen Gu
On 2024/11/23 00:03, Alexandra Winter wrote: > > > On 22.11.24 08:16, Wen Gu wrote: >> We encountered a LGR/link use-after-free issue, which manifested as >> the LGR/link refcnt reaching 0 early and entering the clear process, >> making resource access unsafe. >> >> refcount_t: addition on 0; use-after-free. >> WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140 >> Workqueue: events smc_lgr_terminate_work [smc] >> Call trace: >> refcount_warn_saturate+0x9c/0x140 >> __smc_lgr_terminate.part.45+0x2a8/0x370 [smc] >> smc_lgr_terminate_work+0x28/0x30 [smc] >> process_one_work+0x1b8/0x420 >> worker_thread+0x158/0x510 >> kthread+0x114/0x118 >> >> or >> >> refcount_t: underflow; use-after-free. >> WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140 >> Workqueue: smc_hs_wq smc_listen_work [smc] >> Call trace: >> refcount_warn_saturate+0xf0/0x140 >> smcr_link_put+0x1cc/0x1d8 [smc] >> smc_conn_free+0x110/0x1b0 [smc] >> smc_conn_abort+0x50/0x60 [smc] >> smc_listen_find_device+0x75c/0x790 [smc] >> smc_listen_work+0x368/0x8a0 [smc] >> process_one_work+0x1b8/0x420 >> worker_thread+0x158/0x510 >> kthread+0x114/0x118 >> >> It is caused by repeated release of LGR/link refcnt. One suspect is that >> smc_conn_free() is called repeatedly because some smc_conn_free() are not >> protected by sock lock. >> >> Calls under socklock | Calls not under socklock >> ------------------------------------------------------- >> lock_sock(sk) | smc_conn_abort >> smc_conn_free | \- smc_conn_free >> \- smcr_link_put | \- smcr_link_put (duplicated) >> release_sock(sk) >> >> So make sure smc_conn_free() is called under the sock lock. >> >> Fixes: 8cf3f3e42374 ("net/smc: use helper smc_conn_abort() in listen processing") >> Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> >> Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> >> Co-developed-by: Kai <KaiShen@linux.alibaba.com> >> Signed-off-by: Kai <KaiShen@linux.alibaba.com> >> Signed-off-by: Wen Gu <guwen@linux.alibaba.com> >> --- >> net/smc/af_smc.c | 25 +++++++++++++++++++++---- >> 1 file changed, 21 insertions(+), 4 deletions(-) >> >> diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c >> index ed6d4d520bc7..e0a7a0151b11 100644 >> --- a/net/smc/af_smc.c >> +++ b/net/smc/af_smc.c >> @@ -973,7 +973,8 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, >> return smc_connect_fallback(smc, reason_code); >> } >> >> -static void smc_conn_abort(struct smc_sock *smc, int local_first) >> +static void __smc_conn_abort(struct smc_sock *smc, int local_first, >> + bool locked) >> { >> struct smc_connection *conn = &smc->conn; >> struct smc_link_group *lgr = conn->lgr; >> @@ -982,11 +983,27 @@ static void smc_conn_abort(struct smc_sock *smc, int local_first) >> if (smc_conn_lgr_valid(conn)) >> lgr_valid = true; >> >> - smc_conn_free(conn); >> + if (!locked) { >> + lock_sock(&smc->sk); >> + smc_conn_free(conn); >> + release_sock(&smc->sk); >> + } else { >> + smc_conn_free(conn); >> + } >> if (local_first && lgr_valid) >> smc_lgr_cleanup_early(lgr); >> } >> >> +static void smc_conn_abort(struct smc_sock *smc, int local_first) >> +{ >> + __smc_conn_abort(smc, local_first, false); >> +} >> + >> +static void smc_conn_abort_locked(struct smc_sock *smc, int local_first) >> +{ >> + __smc_conn_abort(smc, local_first, true); >> +} >> + >> /* check if there is a rdma device available for this connection. */ >> /* called for connect and listen */ >> static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) >> @@ -1352,7 +1369,7 @@ static int smc_connect_rdma(struct smc_sock *smc, >> >> return 0; >> connect_abort: >> - smc_conn_abort(smc, ini->first_contact_local); >> + smc_conn_abort_locked(smc, ini->first_contact_local); >> mutex_unlock(&smc_client_lgr_pending); >> smc->connect_nonblock = 0; >> >> @@ -1454,7 +1471,7 @@ static int smc_connect_ism(struct smc_sock *smc, >> >> return 0; >> connect_abort: >> - smc_conn_abort(smc, ini->first_contact_local); >> + smc_conn_abort_locked(smc, ini->first_contact_local); >> mutex_unlock(&smc_server_lgr_pending); >> smc->connect_nonblock = 0; >> > > I wonder if this can deadlock, when you take lock_sock so far down in the callchain. > example: > smc_connect will first take lock_sock(sk) and then mutex_lock(&smc_server_lgr_pending); (e.g. in smc_connect_ism()) > wheras > smc_listen_work() will take mutex_lock(&smc_server_lgr_pending); and then lock_sock(sk) (in your __smc_conn_abort(,,false)) > > I am not sure whether this can be called on the same socket, but it looks suspicious to me. > IMHO this two paths can not occur on the same sk. > > All callers of smc_conn_abort() without socklock seem to originate from smc_listen_work(). > That makes me think whether smc_listen_work() should do lock_sock(sk) on a higher level. > Yes, I also think about this question, I guess it is because the new smc sock will be accepted by userspace only after smc_listen_work() is completed. Before that, no userspace operation occurs synchronously with it, so it is not protected by sock lock. But I am not sure if there are other reasons, so I did not aggressively protect the entire smc_listen_work with sock lock, but chose a conservative approach. > Do you have an example which function could collide with smc_listen_work()? > i.e. have you found a way to reproduce this? > We discovered this during our fault injection testing where the rdma driver was rmmod/insmod sporadically during the nginx/wrk 1K connections test. e.g. __smc_lgr_terminate | smc_listen_decline (caused by rmmod mlx5_ib) | (caused by e.g. reg mr fail) -------------------------------------------------------------- lock_sock | smc_conn_kill | smc_conn_abort \- smc_conn_free | \- smc_conn_free release_sock | > > Are you sure that all callers of smc_conn_free(), that are not smc_conn_abort(), do set the socklock? > It seems to me that the path of smc_conn_kill() is not covered by your solution. > smc_conn_free is called in these places: 1. __smc_release (protected by sock lock) 2. smc_conn_abort (partially protected by sock lock) 3. smc_close_active_abort - smc_release(protected by sock lock) - smc_conn_kill - __smc_lgr_terminate/smc_conn_abort_work(protected by sock lock) 4. smc_close_passive_work (protected by sock lock) So only smc_conn_abort->smc_conn_free is not well protected by sock lock. > > Please excuse, that I am not deeply familiar with this code. > I'm just trying to ask helpful questions. Thanks! :)
On 25.11.24 11:00, Wen Gu wrote: >> I wonder if this can deadlock, when you take lock_sock so far down in the callchain. >> example: >> smc_connect will first take lock_sock(sk) and then mutex_lock(&smc_server_lgr_pending); (e.g. in smc_connect_ism()) >> wheras >> smc_listen_work() will take mutex_lock(&smc_server_lgr_pending); and then lock_sock(sk) (in your __smc_conn_abort(,,false)) >> >> I am not sure whether this can be called on the same socket, but it looks suspicious to me. >> > > IMHO this two paths can not occur on the same sk. > >> >> All callers of smc_conn_abort() without socklock seem to originate from smc_listen_work(). >> That makes me think whether smc_listen_work() should do lock_sock(sk) on a higher level. >> > > Yes, I also think about this question, I guess it is because the new smc sock will be > accepted by userspace only after smc_listen_work() is completed. Before that, no userspace > operation occurs synchronously with it, so it is not protected by sock lock. But I am not > sure if there are other reasons, so I did not aggressively protect the entire smc_listen_work > with sock lock, but chose a conservative approach. > >> Do you have an example which function could collide with smc_listen_work()? >> i.e. have you found a way to reproduce this? >> > > We discovered this during our fault injection testing where the rdma driver was rmmod/insmod > sporadically during the nginx/wrk 1K connections test. > > e.g. > > __smc_lgr_terminate | smc_listen_decline > (caused by rmmod mlx5_ib) | (caused by e.g. reg mr fail) > -------------------------------------------------------------- > lock_sock | > smc_conn_kill | smc_conn_abort > \- smc_conn_free | \- smc_conn_free > release_sock | Thank you for the explanations. So the most suspicious scenario is smc_listen_work() colliding with __smc_lgr_terminate() -> smc_conn_kill() of the conn and smc socket that is just under construction by smc_listen_work() (without socklock). I am wondering, if other parts of smc_listen_work() are allowed to run in parallel with smc_conn_kill() of this smc socket?? My impression would be that the whole smc_listen_work() should be protected against smc_conn_kill(), not only smc_conn_free.
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index ed6d4d520bc7..e0a7a0151b11 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -973,7 +973,8 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, return smc_connect_fallback(smc, reason_code); } -static void smc_conn_abort(struct smc_sock *smc, int local_first) +static void __smc_conn_abort(struct smc_sock *smc, int local_first, + bool locked) { struct smc_connection *conn = &smc->conn; struct smc_link_group *lgr = conn->lgr; @@ -982,11 +983,27 @@ static void smc_conn_abort(struct smc_sock *smc, int local_first) if (smc_conn_lgr_valid(conn)) lgr_valid = true; - smc_conn_free(conn); + if (!locked) { + lock_sock(&smc->sk); + smc_conn_free(conn); + release_sock(&smc->sk); + } else { + smc_conn_free(conn); + } if (local_first && lgr_valid) smc_lgr_cleanup_early(lgr); } +static void smc_conn_abort(struct smc_sock *smc, int local_first) +{ + __smc_conn_abort(smc, local_first, false); +} + +static void smc_conn_abort_locked(struct smc_sock *smc, int local_first) +{ + __smc_conn_abort(smc, local_first, true); +} + /* check if there is a rdma device available for this connection. */ /* called for connect and listen */ static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) @@ -1352,7 +1369,7 @@ static int smc_connect_rdma(struct smc_sock *smc, return 0; connect_abort: - smc_conn_abort(smc, ini->first_contact_local); + smc_conn_abort_locked(smc, ini->first_contact_local); mutex_unlock(&smc_client_lgr_pending); smc->connect_nonblock = 0; @@ -1454,7 +1471,7 @@ static int smc_connect_ism(struct smc_sock *smc, return 0; connect_abort: - smc_conn_abort(smc, ini->first_contact_local); + smc_conn_abort_locked(smc, ini->first_contact_local); mutex_unlock(&smc_server_lgr_pending); smc->connect_nonblock = 0;
We encountered a LGR/link use-after-free issue, which manifested as the LGR/link refcnt reaching 0 early and entering the clear process, making resource access unsafe. refcount_t: addition on 0; use-after-free. WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140 Workqueue: events smc_lgr_terminate_work [smc] Call trace: refcount_warn_saturate+0x9c/0x140 __smc_lgr_terminate.part.45+0x2a8/0x370 [smc] smc_lgr_terminate_work+0x28/0x30 [smc] process_one_work+0x1b8/0x420 worker_thread+0x158/0x510 kthread+0x114/0x118 or refcount_t: underflow; use-after-free. WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140 Workqueue: smc_hs_wq smc_listen_work [smc] Call trace: refcount_warn_saturate+0xf0/0x140 smcr_link_put+0x1cc/0x1d8 [smc] smc_conn_free+0x110/0x1b0 [smc] smc_conn_abort+0x50/0x60 [smc] smc_listen_find_device+0x75c/0x790 [smc] smc_listen_work+0x368/0x8a0 [smc] process_one_work+0x1b8/0x420 worker_thread+0x158/0x510 kthread+0x114/0x118 It is caused by repeated release of LGR/link refcnt. One suspect is that smc_conn_free() is called repeatedly because some smc_conn_free() are not protected by sock lock. Calls under socklock | Calls not under socklock ------------------------------------------------------- lock_sock(sk) | smc_conn_abort smc_conn_free | \- smc_conn_free \- smcr_link_put | \- smcr_link_put (duplicated) release_sock(sk) So make sure smc_conn_free() is called under the sock lock. Fixes: 8cf3f3e42374 ("net/smc: use helper smc_conn_abort() in listen processing") Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com> Co-developed-by: Kai <KaiShen@linux.alibaba.com> Signed-off-by: Kai <KaiShen@linux.alibaba.com> Signed-off-by: Wen Gu <guwen@linux.alibaba.com> --- net/smc/af_smc.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)