linux-kernel - Re: [RFC PATCH v2 2/9] mm: pgtable: Add sysctl to enable COW PTE

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: <7db06e3d-5e00-2b8e-ea46-fbcdb82c4416@nvidia.com>
Date:   Tue, 27 Sep 2022 14:22:27 -0700
From:   John Hubbard <jhubbard@...dia.com>
To:     Chih-En Lin <shiyn.lin@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Qi Zheng <zhengqi.arch@...edance.com>,
        David Hildenbrand <david@...hat.com>,
        Matthew Wilcox <willy@...radead.org>,
        Christophe Leroy <christophe.leroy@...roup.eu>
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        Luis Chamberlain <mcgrof@...nel.org>,
        Kees Cook <keescook@...omium.org>,
        Iurii Zaikin <yzaikin@...gle.com>,
        Vlastimil Babka <vbabka@...e.cz>,
        William Kucharski <william.kucharski@...cle.com>,
        "Kirill A . Shutemov" <kirill.shutemov@...ux.intel.com>,
        Peter Xu <peterx@...hat.com>,
        Suren Baghdasaryan <surenb@...gle.com>,
        Arnd Bergmann <arnd@...db.de>,
        Tong Tiangen <tongtiangen@...wei.com>,
        Pasha Tatashin <pasha.tatashin@...een.com>,
        Li kunyu <kunyu@...china.com>, Nadav Amit <namit@...are.com>,
        Anshuman Khandual <anshuman.khandual@....com>,
        Minchan Kim <minchan@...nel.org>,
        Yang Shi <shy828301@...il.com>, Song Liu <song@...nel.org>,
        Miaohe Lin <linmiaohe@...wei.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
        Andy Lutomirski <luto@...nel.org>,
        Fenghua Yu <fenghua.yu@...el.com>,
        Dinglan Peng <peng301@...due.edu>,
        Pedro Fonseca <pfonseca@...due.edu>,
        Jim Huang <jserv@...s.ncku.edu.tw>,
        Huichun Feng <foxhoundsk.tw@...il.com>
Subject: Re: [RFC PATCH v2 2/9] mm: pgtable: Add sysctl to enable COW PTE

On 9/27/22 09:29, Chih-En Lin wrote:
> Add a new sysctl vm.cow_pte to set MMF_COW_PTE_READY flag for enabling
> copy-on-write (COW) to the PTE page table during the next time of fork.
> 
> Since it has a time gap between using the sysctl to enable the COW PTE
> and doing the fork, we use two states to determine the task that wants
> to do COW PTE or already doing it.
> 
> Signed-off-by: Chih-En Lin <shiyn.lin@...il.com>
> ---
>  include/linux/pgtable.h |  6 ++++++
>  kernel/fork.c           |  5 +++++
>  kernel/sysctl.c         |  8 ++++++++
>  mm/Makefile             |  2 +-
>  mm/cow_pte.c            | 39 +++++++++++++++++++++++++++++++++++++++
>  5 files changed, 59 insertions(+), 1 deletion(-)
>  create mode 100644 mm/cow_pte.c
> 
> diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
> index 014ee8f0fbaab..d03d01aefe989 100644
> --- a/include/linux/pgtable.h
> +++ b/include/linux/pgtable.h
> @@ -937,6 +937,12 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma,
>  	__ptep_modify_prot_commit(vma, addr, ptep, pte);
>  }
>  #endif /* __HAVE_ARCH_PTEP_MODIFY_PROT_TRANSACTION */
> +
> +int cow_pte_handler(struct ctl_table *table, int write, void *buffer,
> +		    size_t *lenp, loff_t *ppos);
> +
> +extern int sysctl_cow_pte_pid;

So are setting a global value, to a single pid?? Only one pid at a time
can be set up? 

I think that tells you already that there is a huge API problem here.

As the other thread with Nadav said, this is not a sysctl. It wants to
be a prctl(), at least the way things look so far.


> +
>  #endif /* CONFIG_MMU */
>  
>  /*
> diff --git a/kernel/fork.c b/kernel/fork.c
> index 8a9e92068b150..6981944a7c6ec 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -2671,6 +2671,11 @@ pid_t kernel_clone(struct kernel_clone_args *args)
>  			trace = 0;
>  	}
>  
> +	if (current->mm && test_bit(MMF_COW_PTE_READY, &current->mm->flags)) {
> +		clear_bit(MMF_COW_PTE_READY, &current->mm->flags);
> +		set_bit(MMF_COW_PTE, &current->mm->flags);
> +	}
> +
>  	p = copy_process(NULL, trace, NUMA_NO_NODE, args);
>  	add_latent_entropy();
>  
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index 205d605cacc5b..c4f54412ae3a9 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -2360,6 +2360,14 @@ static struct ctl_table vm_table[] = {
>  		.mode		= 0644,
>  		.proc_handler	= mmap_min_addr_handler,
>  	},
> +	{
> +		.procname	= "cow_pte",
> +		.data		= &sysctl_cow_pte_pid,
> +		.maxlen		= sizeof(int),
> +		.mode		= 0644,
> +		.proc_handler	= cow_pte_handler,
> +		.extra1		= SYSCTL_ZERO,
> +	},
>  #endif
>  #ifdef CONFIG_NUMA
>  	{
> diff --git a/mm/Makefile b/mm/Makefile
> index 9a564f8364035..7a568d5066ee6 100644
> --- a/mm/Makefile
> +++ b/mm/Makefile
> @@ -40,7 +40,7 @@ mmu-y			:= nommu.o
>  mmu-$(CONFIG_MMU)	:= highmem.o memory.o mincore.o \
>  			   mlock.o mmap.o mmu_gather.o mprotect.o mremap.o \
>  			   msync.o page_vma_mapped.o pagewalk.o \
> -			   pgtable-generic.o rmap.o vmalloc.o
> +			   pgtable-generic.o rmap.o vmalloc.o cow_pte.o
>  
>  
>  ifdef CONFIG_CROSS_MEMORY_ATTACH
> diff --git a/mm/cow_pte.c b/mm/cow_pte.c
> new file mode 100644
> index 0000000000000..4e50aa4294ce7
> --- /dev/null
> +++ b/mm/cow_pte.c
> @@ -0,0 +1,39 @@
> +// SPDX-License-Identifier: GPL-2.0
> +#include <linux/sysctl.h>
> +#include <linux/pgtable.h>
> +#include <linux/sched.h>
> +#include <linux/sched/coredump.h>
> +#include <linux/pid.h>
> +
> +/* sysctl will write to this variable */
> +int sysctl_cow_pte_pid = -1;
> +
> +static void set_cow_pte_task(void)
> +{
> +	struct pid *pid;
> +	struct task_struct *task;
> +
> +	pid = find_get_pid(sysctl_cow_pte_pid);

This seems to be missing a corresponding call to put_pid().

thanks,

-- 
John Hubbard
NVIDIA