linux-kernel - Re: [PATCH 1/1] arm64/sve: Fix wrong free for task->thread.sve

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: <32b59c08-cc59-60d6-16c7-ffb5582c2901@arm.com>
Date:   Fri, 27 Sep 2019 12:04:37 +0100
From:   Julien Grall <julien.grall@....com>
To:     Masayoshi Mizuma <msys.mizuma@...il.com>,
        Catalin Marinas <catalin.marinas@....com>,
        Will Deacon <will@...nel.org>,
        linux-arm-kernel@...ts.infradead.org
Cc:     Masayoshi Mizuma <m.mizuma@...fujitsu.com>,
        linux-kernel@...r.kernel.org, Dave P Martin <Dave.Martin@....com>
Subject: Re: [PATCH 1/1] arm64/sve: Fix wrong free for task->thread.sve_state

(+ Dave)

Hi,

Thank you for the patch.

On 26/09/2019 20:08, Masayoshi Mizuma wrote:
> From: Masayoshi Mizuma <m.mizuma@...fujitsu.com>
> 
> The system which has SVE feature crashed because of
> the memory pointed by task->thread.sve_state was destroyed
> by someone.
> 
> That is because sve_state is freed while the forking the
> child process. The child process has the pointer of sve_state
> which is same as the parent's because the child's task_struct
> is copied from the parent's one. If the copy_process()
> fails as an error on somewhere, for example, copy_creds(),
> then the sve_state is freed even if the parent is alive.
> The flow is as follows.
> 
> copy_process
>          p = dup_task_struct
>              => arch_dup_task_struct
>                  *dst = *src;  // copy the entire region.
> :
>          retval = copy_creds
>          if (retval < 0)
>                  goto bad_fork_free;
> :
> bad_fork_free:
> ...
>          delayed_free_task(p);
>            => free_task
>               => arch_release_task_struct
>                  => fpsimd_release_task
>                     => __sve_free
>                        => kfree(task->thread.sve_state);
>                           // free the parent's sve_state

The flow makes sense to me and I agree you would end up to free the parent's state.

> 
> Add a flag in task->thread which shows the fork is in progress.
> If the fork is in progress, that means the child has the pointer
> to the parent's sve_state, doesn't free the sve_state.

I haven't fully investigate it yet but I was wondering if we could just clear 
task->thread.sve_state for the child in arch_dup_task_struct().

I saw the comment on top of function mentioning potential issue to do it there. 
I understand that you may not be able to clear TIF_SVE in the function, but I 
don't understand why clearing just task->thread.sve_state would be an issue.

The only risk I can see is TIF_SVE may be set with task->thread.sve_state to be 
NULL. But this is a new task, so I don't think there are risk here to have it 
unsync. Dave?

> 
> Signed-off-by: Masayoshi Mizuma <m.mizuma@...fujitsu.com>
> Reported-by: Hidetoshi Seto <seto.hidetoshi@...fujitsu.com>
> ---
>   arch/arm64/include/asm/processor.h | 1 +
>   arch/arm64/kernel/fpsimd.c         | 6 ++++--
>   arch/arm64/kernel/process.c        | 2 ++
>   3 files changed, 7 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/processor.h b/arch/arm64/include/asm/processor.h
> index 5623685c7d13..3ca3e350145a 100644
> --- a/arch/arm64/include/asm/processor.h
> +++ b/arch/arm64/include/asm/processor.h
> @@ -143,6 +143,7 @@ struct thread_struct {
>   	unsigned long		fault_address;	/* fault info */
>   	unsigned long		fault_code;	/* ESR_EL1 value */
>   	struct debug_info	debug;		/* debugging */
> +	unsigned int		fork_in_progress;
>   #ifdef CONFIG_ARM64_PTR_AUTH
>   	struct ptrauth_keys	keys_user;
>   #endif
> diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
> index 37d3912cfe06..8641db4cb062 100644
> --- a/arch/arm64/kernel/fpsimd.c
> +++ b/arch/arm64/kernel/fpsimd.c
> @@ -202,8 +202,10 @@ static bool have_cpu_fpsimd_context(void)
>    */
>   static void __sve_free(struct task_struct *task)
>   {
> -	kfree(task->thread.sve_state);
> -	task->thread.sve_state = NULL;
> +	if (!task->thread.fork_in_progress) {
> +		kfree(task->thread.sve_state);
> +		task->thread.sve_state = NULL;
> +	}
>   }
>   
>   static void sve_free(struct task_struct *task)
> diff --git a/arch/arm64/kernel/process.c b/arch/arm64/kernel/process.c
> index a47462def04b..8ac0ee4e5f76 100644
> --- a/arch/arm64/kernel/process.c
> +++ b/arch/arm64/kernel/process.c
> @@ -347,6 +347,7 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
>   	if (current->mm)
>   		fpsimd_preserve_current_state();
>   	*dst = *src;
> +	dst->thread.fork_in_progress = 1;
>   
>   	return 0;
>   }
> @@ -365,6 +366,7 @@ int copy_thread(unsigned long clone_flags, unsigned long stack_start,
>   	 * and disable discard SVE state for p:
>   	 */
>   	clear_tsk_thread_flag(p, TIF_SVE);
> +	p->thread.fork_in_progress = 0;
>   	p->thread.sve_state = NULL;
>   
>   	/*
> 

Cheers,

-- 
Julien Grall