Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH] powerpc/signal32: Use fault_in_pages_readable() to
 prefault user context
To:     Mathieu Malaterre <malat@debian.org>
Cc:     Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Paul Mackerras <paulus@samba.org>,
        Michael Ellerman <mpe@ellerman.id.au>,
        LKML <linux-kernel@vger.kernel.org>,
        linuxppc-dev <linuxppc-dev@lists.ozlabs.org>
References: <20180424141759.F02AF6C59D@po15720vm.idsi0.si.c-s.fr>
 <CA+7wUsz6NNG=sVUYL1n+ntAJUWNvjtpjCHzJmPuQ4Zb7TuFrZA@mail.gmail.com>
 <CA+7wUsxXwRVHDrE_bzog2+m+kdyDsC4MnjB4_BMc2owDUe7d8A@mail.gmail.com>
From:   Christophe LEROY <christophe.leroy@c-s.fr>
Message-ID: <7d61f98e-8cbb-91cf-97d4-f861758ed11d@c-s.fr>
Date:   Tue, 24 Apr 2018 18:05:12 +0200
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
 Thunderbird/52.7.0
MIME-Version: 1.0
In-Reply-To: <CA+7wUsxXwRVHDrE_bzog2+m+kdyDsC4MnjB4_BMc2owDUe7d8A@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: fr
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


Le 24/04/2018 à 16:50, Mathieu Malaterre a écrit :
> On Tue, Apr 24, 2018 at 4:45 PM, Mathieu Malaterre <malat@debian.org> wrote:
>> On Tue, Apr 24, 2018 at 4:17 PM, Christophe Leroy
>> <christophe.leroy@c-s.fr> wrote:
>>> Use fault_in_pages_readable() to prefault user context
>>> instead of open coding
>>>
>>> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
>>> ---
>>>   arch/powerpc/kernel/signal_32.c | 13 +++++--------
>>>   1 file changed, 5 insertions(+), 8 deletions(-)
>>>
>>> diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
>>> index 492f03451877..cfacb2726152 100644
>>> --- a/arch/powerpc/kernel/signal_32.c
>>> +++ b/arch/powerpc/kernel/signal_32.c
>>> @@ -25,6 +25,7 @@
>>>   #include <linux/errno.h>
>>>   #include <linux/elf.h>
>>>   #include <linux/ptrace.h>
>>> +#include <linux/pagemap.h>
>>>   #include <linux/ratelimit.h>
>>>   #ifdef CONFIG_PPC64
>>>   #include <linux/syscalls.h>
>>> @@ -1045,7 +1046,6 @@ long sys_swapcontext(struct ucontext __user *old_ctx,
>>>                       struct ucontext __user *new_ctx,
>>>                       int ctx_size, int r6, int r7, int r8, struct pt_regs *regs)
>>>   {
>>> -       unsigned char tmp __maybe_unused;
>>>          int ctx_has_vsx_region = 0;
>>>
>>>   #ifdef CONFIG_PPC64
>>> @@ -1109,9 +1109,8 @@ long sys_swapcontext(struct ucontext __user *old_ctx,
>>>          }
>>>          if (new_ctx == NULL)
>>>                  return 0;
>>> -       if (!access_ok(VERIFY_READ, new_ctx, ctx_size)
>>> -           || __get_user(tmp, (u8 __user *) new_ctx)
>>> -           || __get_user(tmp, (u8 __user *) new_ctx + ctx_size - 1))
>>> +       if (!access_ok(VERIFY_READ, new_ctx, ctx_size) ||
>>> +           fault_in_pages_readable((u8 __user *)new_ctx, ctx_size))
>>>                  return -EFAULT;
>>>
>>>          /*
>>> @@ -1231,7 +1230,6 @@ int sys_debug_setcontext(struct ucontext __user *ctx,
>>>   {
>>>          struct sig_dbg_op op;
>>>          int i;
>>> -       unsigned char tmp __maybe_unused;
>>>          unsigned long new_msr = regs->msr;
>>>   #ifdef CONFIG_PPC_ADV_DEBUG_REGS
>>>          unsigned long new_dbcr0 = current->thread.debug.dbcr0;
>>> @@ -1287,9 +1285,8 @@ int sys_debug_setcontext(struct ucontext __user *ctx,
>>>          current->thread.debug.dbcr0 = new_dbcr0;
>>>   #endif
>>>
>>> -       if (!access_ok(VERIFY_READ, ctx, sizeof(*ctx))
>>> -           || __get_user(tmp, (u8 __user *) ctx)
>>> -           || __get_user(tmp, (u8 __user *) (ctx + 1) - 1))
>>> +       if (!access_ok(VERIFY_READ, ctx, sizeof(*ctx)) ||
>>> +           fault_in_pages_readable((u8 __user *)ctx, 1))
>>
>> I believe you meant:
>>
>> fault_in_pages_readable((u8 __user *)new_ctx, ctx_size)
>>
>> Since (u8 __user *) (ctx + 1) - 1 really is (u8 __user *) new_ctx + ctx_size - 1
> 
> Without the copy/paste errors:
> 
> [...]
> fault_in_pages_readable((u8 __user *)ctx, sizeof(*ctx))
> 
> Since (u8 __user *) (ctx + 1) - 1 really is (u8 __user *) ctx + sizeof(*ctx) - 1
> [...]

Oops you're right thanks.
v2 submitted

Christophe


>>>                  return -EFAULT;
>>>
>>>          /*
>>> --
>>> 2.13.3
>>>