diff mbox series

[V3,11/22] bpf: Replace cant_sleep() with cant_migrate()

Message ID	20200224145643.583038889@linutronix.de
State	Accepted
Delegated to:	BPF Maintainers
Headers	show Return-Path: <bpf-owner@vger.kernel.org> Message-Id: <20200224145643.583038889@linutronix.de> User-Agent: quilt/0.65 Date: Mon, 24 Feb 2020 15:01:42 +0100 From: Thomas Gleixner <tglx@linutronix.de> To: LKML <linux-kernel@vger.kernel.org> Cc: David Miller <davem@davemloft.net>, bpf@vger.kernel.org, netdev@vger.kernel.org, Alexei Starovoitov <ast@kernel.org>, Daniel Borkmann <daniel@iogearbox.net>, Sebastian Sewior <bigeasy@linutronix.de>, Peter Zijlstra <peterz@infradead.org>, Clark Williams <williams@redhat.com>, Steven Rostedt <rostedt@goodmis.org>, Juri Lelli <juri.lelli@redhat.com>, Ingo Molnar <mingo@kernel.org>, Mathieu Desnoyers <mathieu.desnoyers@efficios.com>, Vinicius Costa Gomes <vinicius.gomes@intel.com>, Jakub Kicinski <kuba@kernel.org> Subject: [patch V3 11/22] bpf: Replace cant_sleep() with cant_migrate() References: <20200224140131.461979697@linutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Sender: bpf-owner@vger.kernel.org Precedence: bulk
Series	bpf: Make BPF and PREEMPT_RT co-exist \| expand [V3,00/22] bpf: Make BPF and PREEMPT_RT co-exist [V3,01/22] bpf: Tighten the requirements for preallocated hash maps [V3,02/22] bpf: Enforce preallocation for instrumentation programs on RT [V3,03/22] bpf: Update locking comment in hashtab code [V3,04/22] bpf/tracing: Remove redundant preempt_disable() in __bpf_trace_run() [V3,06/22] bpf/trace: Remove redundant preempt_disable from trace_call_bpf() [V3,07/22] perf/bpf: Remove preempt disable around BPF invocation [V3,08/22] bpf: Remove recursion prevention from rcu free callback [V3,09/22] bpf: Dont iterate over possible CPUs with interrupts disabled [V3,10/22] bpf: Provide bpf_prog_run_pin_on_cpu() helper [V3,11/22] bpf: Replace cant_sleep() with cant_migrate() [V3,12/22] bpf: Use bpf_prog_run_pin_on_cpu() at simple call sites. [V3,13/22] bpf/tests: Use migrate disable instead of preempt disable [V3,14/22] bpf: Use migrate_disable/enabe() in trampoline code. [V3,15/22] bpf: Use migrate_disable/enable in array macros and cgroup/lirc code. [V3,16/22] bpf: Provide recursion prevention helpers [V3,17/22] bpf: Use recursion prevention helpers in hashtab code [V3,18/22] bpf: Replace open coded recursion prevention in sys_bpf() [V3,19/22] bpf: Factor out hashtab bucket lock operations [V3,20/22] bpf: Prepare hashtab locking for PREEMPT_RT [V3,21/22] bpf, lpm: Make locking RT friendly [V3,22/22] bpf/stackmap: Dont trylock mmap_sem with PREEMPT_RT and interrupts disabled

Commit Message

Thomas Gleixner Feb. 24, 2020, 2:01 p.m. UTC

As already discussed in the previous change which introduced
BPF_RUN_PROG_PIN_ON_CPU() BPF only requires to disable migration to
guarantee per CPUness.

If RT substitutes the preempt disable based migration protection then the
cant_sleep() check will obviously trigger as preemption is not disabled.

Replace it by cant_migrate() which maps to cant_sleep() on a non RT kernel
and will verify that migration is disabled on a full RT kernel.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/filter.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff mbox series

Patch

--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -561,7 +561,7 @@  DECLARE_STATIC_KEY_FALSE(bpf_stats_enabl
 
 #define __BPF_PROG_RUN(prog, ctx, dfunc)	({			\
 	u32 ret;							\
-	cant_sleep();							\
+	cant_migrate();							\
 	if (static_branch_unlikely(&bpf_stats_enabled_key)) {		\
 		struct bpf_prog_stats *stats;				\
 		u64 start = sched_clock();				\