linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Al Viro <viro@zeniv.linux.org.uk>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	KVM list <kvm@vger.kernel.org>
Subject: Re: [PATCH 8/9] x86: kvm_hv_set_msr(): use __put_user() instead of 32bit __clear_user()
Date: Sat, 30 May 2020 15:31:47 +0100	[thread overview]
Message-ID: <20200530143147.GN23230@ZenIV.linux.org.uk> (raw)
In-Reply-To: <CAHk-=wgq2dzOdN4_=eY-XwxmcgyBM_esnPtXCvz1zStZKjiHKA@mail.gmail.com>

On Fri, May 29, 2020 at 04:52:59PM -0700, Linus Torvalds wrote:
> On Fri, May 29, 2020 at 4:27 PM Al Viro <viro@zeniv.linux.org.uk> wrote:
> > a/arch/x86/kvm/hyperv.c
> > -               if (__clear_user((void __user *)addr, sizeof(u32)))
> > +               if (__put_user(0, (u32 __user *)addr))
> 
> I'm not doubting that this is a correct transformation and an
> improvement, but why is it using that double-underscore version in the
> first place?
> 
> There's a __copy_to_user() in kvm_hv_set_msr_pw() in addition to this
> one in kvm_hv_set_msr(). Both go back to 2011 and commit 8b0cedff040b
> ("KVM: use __copy_to_user/__clear_user to write guest page") and both
> look purely like "pointlessly avoid the access_ok".
> 
> All these KVM "optimizations" seem entirely pointless, since
> access_ok() isn't the problem. And the address _claims_ to be
> verified, but I'm not seeing it. There is not a single 'access_ok()'
> anywhere in arch/x86/kvm/ that I can see.
> 
> It looks like the argument for the address being validated is that it
> comes from "gfn_to_hva()", which should only return
> host-virtual-addresses. That may be true.
> 
> But "should" is not "does", and honestly, the cost of gfn_to_hva() is
> high enough that then using that as an argument for removing
> "access_ok()" smells.
> 
> So I would suggest just removing all these completely bogus
> double-underscore versions. It's pointless, it's wrong, and it's
> unsafe.

It's a bit trickier than that, but I want to deal with that at the same
time as the rest of kvm/vhost stuff.  So for this series I just went
for minimal change.  There's quite a pile of vhost and kvm stuff,
but it's not ready yet - wait for the next cycle.

  reply	other threads:[~2020-05-30 14:31 UTC|newest]

Thread overview: 92+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-28 23:40 [PATCHES] uaccess base Al Viro
2020-05-28 23:48 ` [PATCHES] uaccess readdir Al Viro
2020-05-28 23:49   ` [PATCH 4/6] switch readdir(2) to unsafe_copy_dirent_name() Al Viro
2020-05-28 23:49     ` [PATCH 5/6] readdir.c: get compat_filldir() more or less in sync with filldir() Al Viro
2020-05-28 23:49     ` [PATCH 6/6] readdir.c: get rid of the last __put_user(), drop now-useless access_ok() Al Viro
2020-05-28 23:49 ` [PATCH 1/6] uaccess: Add user_read_access_begin/end and user_write_access_begin/end Al Viro
2020-05-28 23:49   ` [PATCH 2/6] uaccess: Selectively open read or write user access Al Viro
2020-05-28 23:49   ` [PATCH 3/6] drm/i915/gem: Replace user_access_begin by user_write_access_begin Al Viro
2020-05-28 23:57 ` [PATCHES] uaccess __copy_from_user() Al Viro
2020-05-28 23:58   ` [PATCH 1/2] firewire: switch ioctl_queue_iso to use of copy_from_user() Al Viro
2020-05-28 23:58     ` [PATCH 2/2] pstore: switch to copy_from_user() Al Viro
2020-05-29  0:03 ` [PATCHES] uaccess __copy_to_user() Al Viro
2020-05-29  0:04   ` [PATCH 1/2] esas2r: don't bother with __copy_to_user() Al Viro
2020-05-29  0:04     ` [PATCH 2/2] dlmfs: convert dlmfs_file_read() to copy_to_user() Al Viro
2020-05-29  1:27       ` Linus Torvalds
2020-05-29  1:47         ` Al Viro
2020-05-29  1:54           ` Linus Torvalds
2020-05-29  3:10             ` Al Viro
2020-05-29  3:42               ` Linus Torvalds
2020-05-29 20:46                 ` Al Viro
2020-05-29 20:57                   ` Linus Torvalds
2020-05-29 21:06                     ` Al Viro
2020-05-29  0:09 ` [PATCHES] uaccess __put_user() Al Viro
2020-05-29  0:10   ` [PATCH 1/3] compat sysinfo(2): don't bother with field-by-field copyout Al Viro
2020-05-29  0:10     ` [PATCH 2/3] scsi_ioctl.c: switch SCSI_IOCTL_GET_IDLUN to copy_to_user() Al Viro
2020-05-29  0:10     ` [PATCH 3/3] pcm_native: result of put_user() needs to be checked Al Viro
2020-05-29  0:34 ` [PATCHES] uaccess comedi compat Al Viro
2020-05-29  0:35   ` [PATCH 01/10] comedi: move compat ioctl handling to native fops Al Viro
2020-05-29  0:35     ` [PATCH 02/10] comedi: get rid of indirection via translated_ioctl() Al Viro
2020-05-29 10:34       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 03/10] comedi: get rid of compat_alloc_user_space() mess in COMEDI_CHANINFO compat Al Viro
2020-05-29 10:35       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 04/10] comedi: get rid of compat_alloc_user_space() mess in COMEDI_RANGEINFO compat Al Viro
2020-05-29 10:35       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 05/10] comedi: get rid of compat_alloc_user_space() mess in COMEDI_INSN compat Al Viro
2020-05-29 10:05       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 06/10] comedi: get rid of compat_alloc_user_space() mess in COMEDI_INSNLIST compat Al Viro
2020-05-29 10:36       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 07/10] comedi: lift copy_from_user() into callers of __comedi_get_user_cmd() Al Viro
2020-05-29 10:37       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 08/10] comedi: do_cmdtest_ioctl(): lift copyin/copyout into the caller Al Viro
2020-05-29 10:37       ` Ian Abbott
2020-05-29  0:35     ` [PATCH 09/10] comedi: do_cmd_ioctl(): " Al Viro
2020-05-29 10:38       ` Ian Abbott
2020-05-29 10:34     ` [PATCH 01/10] comedi: move compat ioctl handling to native fops Ian Abbott
2020-05-29 10:48   ` [PATCHES] uaccess comedi compat Ian Abbott
2020-05-29 14:15     ` Al Viro
2020-05-29  0:40 ` [PATCHES] uaccess i915 Al Viro
2020-05-29  5:06   ` Jani Nikula
2020-05-29 14:17     ` Al Viro
2020-05-29  0:41 ` [PATCH 1/5] i915: switch query_{topology,engine}_info() to copy_to_user() Al Viro
2020-05-29  0:41   ` [PATCH 2/5] i915: switch copy_perf_config_registers_or_number() to unsafe_put_user() Al Viro
2020-05-29  0:41   ` [PATCH 3/5] i915 compat ioctl(): just use drm_ioctl_kernel() Al Viro
2020-05-29  0:41   ` [PATCH 4/5] i915: alloc_oa_regs(): get rid of pointless access_ok() Al Viro
2020-05-29  0:41   ` [PATCH 5/5] i915:get_engines(): " Al Viro
2020-05-29 23:26 ` [PATCHES] uaccess misc Al Viro
2020-05-29 23:54   ` Linus Torvalds
2020-05-29 23:57     ` Linus Torvalds
2020-05-29 23:27 ` [PATCH 1/9] pselect6() and friends: take handling the combined 6th/7th args into helper Al Viro
2020-05-29 23:27   ` [PATCH 2/9] binfmt_elf: don't bother with __{put,copy_to}_user() Al Viro
2020-05-29 23:27   ` [PATCH 3/9] binfmt_elf_fdpic: don't use __... uaccess primitives Al Viro
2020-05-29 23:27   ` [PATCH 4/9] binfmt_flat: don't use __put_user() Al Viro
2020-05-29 23:27   ` [PATCH 5/9] x86: switch cp_stat64() to unsafe_put_user() Al Viro
2020-05-29 23:27   ` [PATCH 6/9] TEST_ACCESS_OK _never_ had been checked anywhere Al Viro
2020-05-29 23:27   ` [PATCH 7/9] user_regset_copyout_zero(): use clear_user() Al Viro
2020-05-29 23:27   ` [PATCH 8/9] x86: kvm_hv_set_msr(): use __put_user() instead of 32bit __clear_user() Al Viro
2020-05-29 23:52     ` Linus Torvalds
2020-05-30 14:31       ` Al Viro [this message]
2020-05-30 14:52         ` Al Viro
2020-05-30 16:20         ` Paolo Bonzini
2020-05-30 17:57           ` Linus Torvalds
2020-05-30 18:38             ` Al Viro
2020-05-30 18:52               ` Linus Torvalds
2020-05-30 19:14                 ` Al Viro
2020-05-30 19:20                   ` Linus Torvalds
2020-05-30 19:42                     ` Al Viro
2020-05-30 20:43                       ` Al Viro
2020-05-30 19:19                 ` Al Viro
2020-05-30 19:27                   ` Al Viro
2020-05-29 23:28 ` [PATCH 9/9] bpf: make bpf_check_uarg_tail_zero() use check_zeroed_user() Al Viro
2020-05-31 16:35   ` Alexei Starovoitov
2020-05-29 23:39 ` [PATCHES] uaccess hpsa Al Viro
2020-05-29 23:40   ` [PATCH 1/4] hpsa passthrough: lift {BIG_,}IOCTL_Command_struct copy{in,out} into hpsa_ioctl() Al Viro
2020-05-29 23:40     ` [PATCH 2/4] hpsa: don't bother with vmalloc for BIG_IOCTL_Command_struct Al Viro
2020-05-29 23:40     ` [PATCH 3/4] hpsa: get rid of compat_alloc_user_space() Al Viro
2020-05-29 23:40     ` [PATCH 4/4] hpsa_ioctl(): tidy up a bit Al Viro
2020-06-03  1:57   ` [PATCHES] uaccess hpsa Martin K. Petersen
2020-06-03 18:37   ` Don.Brace
2020-06-03 19:17     ` Al Viro
2020-06-03 20:53       ` Martin K. Petersen
2020-06-03 20:54         ` Al Viro
2020-06-04 14:18           ` Martin K. Petersen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200530143147.GN23230@ZenIV.linux.org.uk \
    --to=viro@zeniv.linux.org.uk \
    --cc=kvm@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.com \
    --cc=torvalds@linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).