From: Zack Weinberg <zackw@panix.com>
To: Florian Weimer <fweimer@redhat.com>
Cc: GNU C Library <libc-alpha@sourceware.org>
Subject: Re: [PATCH v7] getrandom system call wrapper [BZ #17252]
Date: Wed, 16 Nov 2016 16:41:00 -0000 [thread overview]
Message-ID: <CAKCAbMgkFGz_D1jeyjb-2em1mdxJxc+3KXr3miux6O_uBtJHiQ@mail.gmail.com> (raw)
In-Reply-To: <c30852c5-9a52-83ea-4af1-310f97febccd@redhat.com>
On Wed, Nov 16, 2016 at 10:52 AM, Florian Weimer <fweimer@redhat.com> wrote:
> On 11/16/2016 04:20 PM, Zack Weinberg wrote:
>> I am seriously considering escalating my disagreement here to a
>> formal objection. I would like to know why you think it is
>> NECESSARY - not merely convenient or consistent with other stuff -
>> for this function to be a cancellation point.
>
> It's necessary if you ever want to cancel a hanging getrandom in a
> context where you cannot install a signal handler (so that you can
> trigger EINTR when getrandom is stuck).
That only pushes the question back a level. When would it ever be
necessary to do that? Be as concrete as you possibly can. Actual
code from a real program, if possible.
> I really don't understand why cancellation points are widely
> considered as evil. Most code does not use cancellation in a
> correct way, and it will not improve if we simply stop adding new
> cancellation points.
I would argue that most code does not use cancellation correctly in
large part _because_ the set of cancellation points is so large and
amorphous. It is the same problem that people have with exceptions in
C++; because practically everything you might do exposes you to
cancellation, it's so difficult to know how to write cancel-safe code
that people just apply the big hammer of not using it at all. Another
useful example is multithreading versus coroutines with explicit yield
points---the latter can be less efficient and/or more verbose, but
it's so much easier to debug that it's worth it.
There's not a lot we can do about the cancellation-point set in POSIX
being so large, but we can at least not make things worse, by not
adding additional cancellation points.
> at least in library code, it is impossible to introduce cancellation
> into a system call where the wrapper does not support it (because
> you cannot fake your own version of cancellation with a do-nothing
> signal handler).
From the perspective that as few operations as possible should be
cancellation points, this is a Good Thing. And I don't see why it
would be a problem for getrandom in particular.
zw
next prev parent reply other threads:[~2016-11-16 16:41 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-10 21:03 [PATCH] Add getrandom implementation " Florian Weimer
2016-06-10 21:31 ` Joseph Myers
2016-06-10 21:36 ` Joseph Myers
2016-06-10 22:00 ` Paul Eggert
2016-06-10 22:06 ` Joseph Myers
2016-06-11 11:13 ` Florian Weimer
2016-06-11 20:10 ` Paul Eggert
2016-06-10 22:15 ` Roland McGrath
2016-06-10 22:40 ` Joseph Myers
2016-06-10 22:45 ` Roland McGrath
2016-06-23 17:21 ` Florian Weimer
2016-06-25 21:58 ` Paul Eggert
2016-09-02 22:23 ` Roland McGrath
2016-06-27 15:07 ` [PATCH v2] " Florian Weimer
2016-06-30 9:33 ` Rical Jasan
2016-09-08 9:53 ` Florian Weimer
2016-09-08 10:13 ` Andreas Schwab
2016-09-08 10:28 ` Florian Weimer
2016-09-08 11:58 ` Rical Jasan
2016-09-08 12:36 ` Florian Weimer
2016-06-30 12:03 ` Zack Weinberg
2016-07-13 13:10 ` Nikos Mavrogiannopoulos
2016-11-14 17:45 ` [PATCH v7] getrandom system call wrapper " Florian Weimer
2016-11-14 18:29 ` Zack Weinberg
2016-11-15 20:57 ` Richard Henderson
2016-11-16 15:11 ` Florian Weimer
2016-11-16 15:20 ` Zack Weinberg
2016-11-16 15:52 ` Florian Weimer
2016-11-16 16:41 ` Zack Weinberg [this message]
2016-11-17 13:02 ` Florian Weimer
2016-11-17 13:46 ` Zack Weinberg
2016-11-17 13:50 ` Florian Weimer
2016-11-17 13:56 ` Zack Weinberg
2016-11-17 15:24 ` Florian Weimer
2016-11-17 17:16 ` Zack Weinberg
2016-11-18 10:27 ` Szabolcs Nagy
2016-11-18 15:46 ` Torvald Riegel
2016-11-18 18:50 ` Zack Weinberg
2016-11-21 16:57 ` Torvald Riegel
2016-11-21 17:12 ` Zack Weinberg
2016-11-21 17:30 ` Torvald Riegel
2016-11-21 17:34 ` Florian Weimer
2016-11-29 8:24 ` Florian Weimer
2016-11-16 18:02 ` Torvald Riegel
2016-11-16 19:53 ` Adhemerval Zanella
2016-11-17 12:52 ` Torvald Riegel
2016-11-18 8:28 ` Szabolcs Nagy
2016-11-18 14:21 ` Torvald Riegel
2016-11-18 15:13 ` Florian Weimer
2016-11-18 16:04 ` Torvald Riegel
2016-11-29 8:16 ` Florian Weimer
2016-11-29 13:56 ` Torvald Riegel
2016-11-29 14:40 ` Florian Weimer
2016-11-29 15:23 ` Torvald Riegel
2016-11-29 15:32 ` Florian Weimer
2016-11-29 15:54 ` Zack Weinberg
2016-11-29 17:53 ` Paul Eggert
2016-11-29 18:11 ` Florian Weimer
2016-11-29 19:37 ` Paul Eggert
2016-11-30 6:09 ` Florian Weimer
2016-11-17 6:21 ` Mike Frysinger
2016-11-18 13:21 ` Florian Weimer
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=CAKCAbMgkFGz_D1jeyjb-2em1mdxJxc+3KXr3miux6O_uBtJHiQ@mail.gmail.com \
--to=zackw@panix.com \
--cc=fweimer@redhat.com \
--cc=libc-alpha@sourceware.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 read-only IMAP folder(s) and NNTP newsgroup(s).