| Message ID | 20220723015809.11553-1-duoming@zju.edu.cn (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | b89fc26f741d9f9efb51cba3e9b241cf1380ec5a |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net] sctp: fix sleep in atomic context bug in timer handlers | expand |
Hi, On Sat, Jul 23, 2022 at 09:58:09AM +0800, Duoming Zhou wrote: > There are sleep in atomic context bugs in timer handlers of sctp > such as sctp_generate_t3_rtx_event(), sctp_generate_probe_event(), > sctp_generate_t1_init_event(), sctp_generate_timeout_event(), > sctp_generate_t3_rtx_event() and so on. > > The root cause is sctp_sched_prio_init_sid() with GFP_KERNEL parameter > that may sleep could be called by different timer handlers which is in > interrupt context. > > One of the call paths that could trigger bug is shown below: > > (interrupt context) > sctp_generate_probe_event > sctp_do_sm > sctp_side_effects > sctp_cmd_interpreter > sctp_outq_teardown > sctp_outq_init This sequence is odd but it is used when handling dup cookies. It tears down whatever was in there and re-inits it. With that, Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> > sctp_sched_set_sched > n->init_sid(..,GFP_KERNEL) > sctp_sched_prio_init_sid //may sleep
Hello: This patch was applied to netdev/net.git (master) by Jakub Kicinski <kuba@kernel.org>: On Sat, 23 Jul 2022 09:58:09 +0800 you wrote: > There are sleep in atomic context bugs in timer handlers of sctp > such as sctp_generate_t3_rtx_event(), sctp_generate_probe_event(), > sctp_generate_t1_init_event(), sctp_generate_timeout_event(), > sctp_generate_t3_rtx_event() and so on. > > The root cause is sctp_sched_prio_init_sid() with GFP_KERNEL parameter > that may sleep could be called by different timer handlers which is in > interrupt context. > > [...] Here is the summary with links: - [net] sctp: fix sleep in atomic context bug in timer handlers https://git.kernel.org/netdev/net/c/b89fc26f741d You are awesome, thank you!
diff --git a/net/sctp/stream_sched.c b/net/sctp/stream_sched.c index 518b1b9bf89..1ad565ed562 100644 --- a/net/sctp/stream_sched.c +++ b/net/sctp/stream_sched.c @@ -160,7 +160,7 @@ int sctp_sched_set_sched(struct sctp_association *asoc, if (!SCTP_SO(&asoc->stream, i)->ext) continue; - ret = n->init_sid(&asoc->stream, i, GFP_KERNEL); + ret = n->init_sid(&asoc->stream, i, GFP_ATOMIC); if (ret) goto err; }
There are sleep in atomic context bugs in timer handlers of sctp such as sctp_generate_t3_rtx_event(), sctp_generate_probe_event(), sctp_generate_t1_init_event(), sctp_generate_timeout_event(), sctp_generate_t3_rtx_event() and so on. The root cause is sctp_sched_prio_init_sid() with GFP_KERNEL parameter that may sleep could be called by different timer handlers which is in interrupt context. One of the call paths that could trigger bug is shown below: (interrupt context) sctp_generate_probe_event sctp_do_sm sctp_side_effects sctp_cmd_interpreter sctp_outq_teardown sctp_outq_init sctp_sched_set_sched n->init_sid(..,GFP_KERNEL) sctp_sched_prio_init_sid //may sleep This patch changes gfp_t parameter of init_sid in sctp_sched_set_sched() from GFP_KERNEL to GFP_ATOMIC in order to prevent sleep in atomic context bugs. Fixes: 5bbbbe32a431 ("sctp: introduce stream scheduler foundations") Signed-off-by: Duoming Zhou <duoming@zju.edu.cn> --- net/sctp/stream_sched.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)