linux-kernel - Re: 6.11/regression/bisected - The commit c1385c1f0ba3 caused a new possible recursive locking detected warning at computer boot.

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] [day] [month] [year] [list]

Message-ID: <2dd6b8e1-6f06-4ab8-b610-c0bd95a397e8@redhat.com>
Date: Sat, 3 Aug 2024 17:48:18 +0200
From: Hans de Goede <hdegoede@...hat.com>
To: Thomas Gleixner <tglx@...utronix.de>,
 Jonathan Cameron <Jonathan.Cameron@...wei.com>,
 Mikhail Gavrilov <mikhail.v.gavrilov@...il.com>, linuxarm@...wei.com
Cc: rafael.j.wysocki@...el.com, guohanjun@...wei.com, gshan@...hat.com,
 miguel.luis@...cle.com, catalin.marinas@....com,
 Linux List Kernel Mailing <linux-kernel@...r.kernel.org>,
 Linux regressions mailing list <regressions@...ts.linux.dev>,
 Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
 Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
 "H. Peter Anvin" <hpa@...or.com>, "Bowman, Terry" <Terry.bowman@....com>
Subject: Re: 6.11/regression/bisected - The commit c1385c1f0ba3 caused a new
 possible recursive locking detected warning at computer boot.

Hi,

On 7/26/24 6:26 PM, Thomas Gleixner wrote:
> On Thu, Jul 25 2024 at 18:13, Jonathan Cameron wrote:
>> On Tue, 23 Jul 2024 18:20:06 +0100
>> Jonathan Cameron <Jonathan.Cameron@...wei.com> wrote:
>>
>>>> This is an interesting corner and perhaps reflects a flawed
>>>> assumption we were making that for this path anything that can happen for an
>>>> initially present CPU can also happen for a hotplugged one. On the hotplugged
>>>> path the lock was always held and hence the static_key_enable() would
>>>> have failed.
> 
> No. The original code invoked this without cpus read locked via:
> 
> acpi_processor_driver.probe()
>    __acpi_processor_start()
>        ....
> 
> and the cpu hotplug callback finds it already set up, so it won't reach
> the static_key_enable() anymore.
> 
>> One bit I need to check out tomorrow is to make sure this doesn't race with the
>> workfn that is used to tear down the same static key on error.
> 
> There is a simpler solution for that. See the uncompiled below.

I was seeing the same lockdep issue as the original reporter on my
AMD Ryzen 9 5950X machine and I can confirm that the suggested patch
(below) fixes this:

Tested-by: Hans de Goede <hdegoede@...hat.com>

Regards,

Hans




> ---
> diff --git a/arch/x86/kernel/cpu/aperfmperf.c b/arch/x86/kernel/cpu/aperfmperf.c
> index b3fa61d45352..0b69bfbf345d 100644
> --- a/arch/x86/kernel/cpu/aperfmperf.c
> +++ b/arch/x86/kernel/cpu/aperfmperf.c
> @@ -306,7 +306,7 @@ static void freq_invariance_enable(void)
>  		WARN_ON_ONCE(1);
>  		return;
>  	}
> -	static_branch_enable(&arch_scale_freq_key);
> +	static_branch_enable_cpuslocked(&arch_scale_freq_key);
>  	register_freq_invariance_syscore_ops();
>  	pr_info("Estimated ratio of average max frequency by base frequency (times 1024): %llu\n", arch_max_freq_ratio);
>  }
> @@ -323,8 +323,10 @@ static void __init bp_init_freq_invariance(void)
>  	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
>  		return;
>  
> -	if (intel_set_max_freq_ratio())
> +	if (intel_set_max_freq_ratio()) {
> +		guard(cpus_read_lock)();
>  		freq_invariance_enable();
> +	}
>  }
>  
>  static void disable_freq_invariance_workfn(struct work_struct *work)
> 
> 
>