driverdev-devel.linuxdriverproject.org archive mirror
 help / color / mirror / Atom feed
From: Al Viro <viro@zeniv.linux.org.uk>
To: Christian Brauner <christian.brauner@ubuntu.com>
Cc: devel@driverdev.osuosl.org, tkjos@android.com,
	keescook@chromium.org, Greg KH <gregkh@linuxfoundation.org>,
	jasowang@redhat.com, linux-fsdevel@vger.kernel.org,
	sargun@sargun.me, hch@infradead.org,
	Xie Yongji <xieyongji@bytedance.com>,
	arve@android.com, joel@joelfernandes.org, hridya@google.com,
	maco@android.com, surenb@google.com
Subject: Re: [PATCH 2/2] binder: Use receive_fd() to receive file from another process
Date: Fri, 16 Apr 2021 17:30:41 +0000	[thread overview]
Message-ID: <YHnJwRvUhaK3IM0l@zeniv-ca.linux.org.uk> (raw)
In-Reply-To: <20210416155815.ayjpnx37dv3a4jos@wittgenstein>

On Fri, Apr 16, 2021 at 05:58:15PM +0200, Christian Brauner wrote:

> They could probably refactor this but I'm not sure why they'd bother. If
> they fail processing any of those files they end up aborting the
> whole transaction.
> (And the original code didn't check the error code btw.)

Wait a sec...  What does aborting the transaction do to descriptor table?
<rereads>
Oh, lovely...

binder_apply_fd_fixups() is deeply misguided.  What it should do is
	* go through t->fd_fixups, reserving descriptor numbers and
putting them into t->buffer (and I'd probably duplicate them into
struct binder_txn_fd_fixup).  Cleanup in case of failure: go through
the list, releasing the descriptors we'd already reserved, doing
fput() on fixup->file in all entries and freeing the entries as
we go.
	* On success, go through the list, doing fd_install() and
freeing the entries.

That's it.  No rereading from the buffer, no binder_deferred_fd_close()
crap, etc.

Again, YOU CAN NOT UNDO fd_install().  Ever.  Kernel can not decide it
shouldn't have put something in descriptor table and take it back.
You can't unring that bell.
_______________________________________________
devel mailing list
devel@linuxdriverproject.org
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel

  parent reply	other threads:[~2021-04-16 17:31 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01  9:09 [PATCH 0/2] Export receive_fd() to modules and do some cleanups Xie Yongji
2021-04-01  9:09 ` [PATCH 1/2] file: Export receive_fd() to modules Xie Yongji
2021-04-01  9:52   ` Greg KH
2021-04-01 10:24     ` Yongji Xie
2021-04-01  9:09 ` [PATCH 2/2] binder: Use receive_fd() to receive file from another process Xie Yongji
2021-04-01  9:54   ` Greg KH
2021-04-01 10:12     ` Yongji Xie
2021-04-01 10:42       ` Greg KH
2021-04-01 11:29         ` Yongji Xie
2021-04-01 11:33           ` Greg KH
2021-04-01 12:28             ` Yongji Xie
2021-04-01 14:09               ` Greg KH
2021-04-02  9:12                 ` Kees Cook
2021-04-01 10:40     ` Christian Brauner
2021-04-01 11:11       ` Yongji Xie
2021-04-16  5:19       ` Al Viro
2021-04-16  5:55         ` Al Viro
2021-04-16 13:42           ` Christian Brauner
2021-04-16 14:09             ` Al Viro
2021-04-16 15:13               ` Christian Brauner
2021-04-16 15:35                 ` Al Viro
2021-04-16 15:58                   ` Christian Brauner
2021-04-16 16:00                     ` Christian Brauner
2021-04-16 17:00                       ` Al Viro
2021-04-16 17:30                     ` Al Viro [this message]
2021-04-17  1:30                       ` Al Viro
2021-04-01  9:53 ` [PATCH 0/2] Export receive_fd() to modules and do some cleanups Greg KH
2021-04-01 10:00   ` Yongji Xie
2021-04-01 10:20 ` Christian Brauner

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=YHnJwRvUhaK3IM0l@zeniv-ca.linux.org.uk \
    --to=viro@zeniv.linux.org.uk \
    --cc=arve@android.com \
    --cc=christian.brauner@ubuntu.com \
    --cc=devel@driverdev.osuosl.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=hch@infradead.org \
    --cc=hridya@google.com \
    --cc=jasowang@redhat.com \
    --cc=joel@joelfernandes.org \
    --cc=keescook@chromium.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=maco@android.com \
    --cc=sargun@sargun.me \
    --cc=surenb@google.com \
    --cc=tkjos@android.com \
    --cc=xieyongji@bytedance.com \
    /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).