lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20220313055512.248571-3-leo.yan@linaro.org>
Date:   Sun, 13 Mar 2022 13:55:11 +0800
From:   Leo Yan <leo.yan@...aro.org>
To:     Sudeep Holla <sudeep.holla@....com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Bryan O'Donoghue <bryan.odonoghue@...aro.org>,
        linux-kernel@...r.kernel.org
Cc:     Leo Yan <leo.yan@...aro.org>
Subject: [PATCH v1 2/3] arch_topology: Handle inconsistent binding of CPU raw capacity

There have a corner case is if we use below DT binding:

  cpu0: cpu@0 {
  	compatible = "arm,cortex-a53";
  	device_type = "cpu";
  	reg = <0x0 0x0>;
  	enable-method = "psci";
  };

  cpu1: cpu@1 {
  	compatible = "arm,cortex-a53";
  	device_type = "cpu";
  	reg = <0x0 0x1>;
  	enable-method = "psci";
  	capacity-dmips-mhz = <1024>;
  };

In this case, CPU0 node misses to bind "capacity-dmips-mhz" property,
and CPU1 node binds the property, this means the CPU raw capacity
binding is inconsistent across all CPUs.

This patch introduces an extra flag 'cap_property_miss' to indicate that
any previous CPU nodes miss binding for "capacity-dmips-mhz" property,
and any new CPU node contains the property, it detects the inconsistent
binding, and sets 'cap_parsing_failed' to true and frees raw capacity
array.

Signed-off-by: Leo Yan <leo.yan@...aro.org>
---
 drivers/base/arch_topology.c | 25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)

diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
index b81777ae6425..0687576e880b 100644
--- a/drivers/base/arch_topology.c
+++ b/drivers/base/arch_topology.c
@@ -294,6 +294,7 @@ bool __init topology_parse_cpu_capacity(struct device_node *cpu_node, int cpu)
 	struct clk *cpu_clk;
 	int ret;
 	u32 cpu_capacity;
+	static bool cap_property_miss;
 
 	if (cap_parsing_failed)
 		return false;
@@ -301,6 +302,20 @@ bool __init topology_parse_cpu_capacity(struct device_node *cpu_node, int cpu)
 	ret = of_property_read_u32(cpu_node, "capacity-dmips-mhz",
 				   &cpu_capacity);
 	if (!ret) {
+		/*
+		 * A previous CPU node misses binding for CPU raw capacity and
+		 * current CPU node finds its property "capacity-dmips-mhz",
+		 * thus the DT binding for "capacity-dmips-mhz" is inconsistent
+		 * across all CPUs.  Set 'cap_parsing_failed' flag and drop the
+		 * CPU raw capacity values.
+		 */
+		if (cap_property_miss) {
+			pr_err("cpu_capacity: binding %pOF raw capacity\n",
+				cpu_node);
+			pr_err("cpu_capacity: partial information: fallback to 1024 for all CPUs\n");
+			goto parsing_failure;
+		}
+
 		if (!raw_capacity) {
 			raw_capacity = kcalloc(num_possible_cpus(),
 					       sizeof(*raw_capacity),
@@ -331,12 +346,18 @@ bool __init topology_parse_cpu_capacity(struct device_node *cpu_node, int cpu)
 			pr_err("cpu_capacity: missing %pOF raw capacity\n",
 				cpu_node);
 			pr_err("cpu_capacity: partial information: fallback to 1024 for all CPUs\n");
-			cap_parsing_failed = true;
-			free_raw_capacity();
+			goto parsing_failure;
+		} else {
+			cap_property_miss = true;
 		}
 	}
 
 	return !ret;
+
+parsing_failure:
+	cap_parsing_failed = true;
+	free_raw_capacity();
+	return !ret;
 }
 
 #ifdef CONFIG_CPU_FREQ
-- 
2.25.1

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ