linux-kernel - Re: [PATCH v2] sched/isolation: Allow "isolcpus=" to skip unknown sub-parameters

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: <20200403011559.GH103677@xz-x1>
Date:   Thu, 2 Apr 2020 21:15:59 -0400
From:   Peter Xu <peterx@...hat.com>
To:     linux-kernel@...r.kernel.org
Cc:     Ming Lei <ming.lei@...hat.com>, Ingo Molnar <mingo@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Juri Lelli <juri.lelli@...hat.com>,
        Luiz Capitulino <lcapitulino@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH v2] sched/isolation: Allow "isolcpus=" to skip unknown
 sub-parameters

On Thu, Apr 02, 2020 at 10:59:29AM -0400, Peter Xu wrote:
> The "isolcpus=" parameter allows sub-parameters to exist before the
> cpulist is specified, and if it sees unknown sub-parameters the whole
> parameter will be ignored.  This design is incompatible with itself
> when we add more sub-parameters to "isolcpus=", because the old
> kernels will not recognize the new "isolcpus=" sub-parameters, then it
> will invalidate the whole parameter so the CPU isolation will not
> really take effect if we start to use the new sub-parameters while
> later we reboot into an old kernel. Instead we will see this when
> booting the old kernel:
> 
>     isolcpus: Error, unknown flag
> 
> The better and compatible way is to allow "isolcpus=" to skip unknown
> sub-parameters, so that even if we add new sub-parameters to it the
> old kernel will still be able to behave as usual even if with the new
> sub-parameter is specified.
> 
> Ideally this patch should be there when we introduce the first
> sub-parameter for "isolcpus=", so it's already a bit late.  However
> late is better than nothing.
> 
> CC: Ming Lei <ming.lei@...hat.com>
> CC: Ingo Molnar <mingo@...hat.com>
> CC: Peter Zijlstra <peterz@...radead.org>
> CC: Juri Lelli <juri.lelli@...hat.com>
> CC: Luiz Capitulino <lcapitulino@...hat.com>
> Suggested-by: Thomas Gleixner <tglx@...utronix.de>
> Signed-off-by: Peter Xu <peterx@...hat.com>
> ---
> v2:
> - only allow isalpha() for sub-parameters [tglx]
> ---
>  kernel/sched/isolation.c | 20 ++++++++++++++++++--
>  1 file changed, 18 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c
> index 008d6ac2342b..c2e8b4a778d6 100644
> --- a/kernel/sched/isolation.c
> +++ b/kernel/sched/isolation.c
> @@ -149,6 +149,9 @@ __setup("nohz_full=", housekeeping_nohz_full_setup);
>  static int __init housekeeping_isolcpus_setup(char *str)
>  {
>  	unsigned int flags = 0;
> +	char *par;
> +	int len;
> +	bool illegal = false;
>  
>  	while (isalpha(*str)) {
>  		if (!strncmp(str, "nohz,", 5)) {
> @@ -169,8 +172,21 @@ static int __init housekeeping_isolcpus_setup(char *str)
>  			continue;
>  		}
>  
> -		pr_warn("isolcpus: Error, unknown flag\n");
> -		return 0;
> +		/*
> +		 * Skip unknown sub-parameter and validate that it is not
> +		 * containing an invalid character.
> +		 */
> +		for (par = str, len = 0; *str && *str != ','; str++, len++)
> +			if (!isalpha(*str))
> +				illegal = true;
> +
> +		if (illegal) {
> +			pr_warn("isolcpus: Invalid flag %.*s\n", len, par);
> +			return 0;
> +		}
> +
> +		pr_info("isolcpus: Skipped unknown flag %.*s\n", len, par);
> +		str++;
>  	}

I just noticed this is still problematic if we want to mark this as
stable, because "managed_irq" violate the "isalpha()" rule already...
It means even if we apply this patch to the stable trees it'll still
think managed_irq as illegal and ignore the whole isolcpus=.

Thomas, do you want me to repost a v3 as v1 plus some pr_warn()s?

Thanks,

-- 
Peter Xu