[Toybox] freadahead()

[Rob Landley]

On 4/11/22 19:03, enh wrote:
> On Mon, Apr 11, 2022 at 3:19 PM Rob Landley <rob at landley.net
> <mailto:rob at landley.net>> wrote:
>     Anyway, if musl and bionic have __freadahead(), I can either #ifdef the glibc
>     hack for it in portability.h, or might just move xgetline() into portability.c
>     and have it do get_line() on systems that don't have __freadahead(). 
> 
> you're still screwed on macOS? (oh, except macOS seems to be just the regular
> BSD header, with everything public and "roughly what you'd expect", so that's
> easily done manually too.)

There's a bunch of stuff in the m4 and glib #ifdefs that show how to do it for
other libraries, although I only really care about 3 and can fall back to
get_line() for the others in portability.c. :)

But part of my thinking is it's easier to convince the posix guys to standardize
the wrapper if the wrapper is more widely implemented. It's already in musl,
Dragonfly BSD, Z/OS...

> yeah, iirc i added everything that glibc had but didn't even look at musl. (if
> you'd asked me even this morning, i'd have told you musl was a strict subset of
> glibc. TIL.)

There are a few exceptions.

>     Honestly, I
>     have no sympathy for the glibc guys who implemented everything EXCEPT
>     __freadahead() because they want to force people to use
>     gnugnugnustallmanftaghnlib:
> 
> 
> citation needed :-)

If you want a list of what they DID implement, the man page I linked to lists
them. (It's one of them group man pages for a bunch of related features.)

If you want examples of gnu pressuring people to use their libraries to spread
their license... they've been pretty open about it? Their darling used to be
libreadline before that got cloned...

https://www.gnu.org/licenses/why-not-lgpl.en.html
https://news.ycombinator.com/item?id=20772053#:~:text=embraced

There are more explicit on-topic references, but... sigh, stuff wanders by over
the years that can be a bit hard to dig up again. They've been at this FOREVER...

https://postgrespro.com/list/thread-id/1596439

That whole
https://www.computerworld.com/article/2732025/gpl-enforcement-sparks-community-flames.html
thing was not my first licensing rodeo. I first encountered the GPL via the EMX
project back in college like a decade before that:

  https://en.wikipedia.org/wiki/EMX_(programming_environment)

My first serious attempt to educate myself on IP law was over 20 years ago now,
sometime before... Grrr, TMF keeps moving their archives around. Lemme splice
together the week with my own archives...

https://landley.net/writing/mirror/fool/rulemaker000501.htm
https://www.fool.com/archive/portfolios/rulemaker/2000/05/02/get-your-copyrights-here.aspx
https://www.fool.com/archive/portfolios/rulemaker/2000/05/03/you-gonna-patent-that.aspx
http://landley.net/writing/mirror/fool/rulemaker000504.htm
http://landley.net/writing/mirror/fool/rulemaker000505.htm

I.E. sometime before that.

I even drove to Boston to talk to Richard Stallman at his office in February
2001. (I'd already mailed him a VHS copy of the musical 1776, so I could explain
he was John Adams and needed ESR's Franklin and Torvalds' Jefferson to get his
agenda through. He didn't take that well, he thought he was Franklin. Sigh.)

Anyway, I was very interested in the IP considerations and ramifications of all
this stuff because licensing and development model were a big part what made
Linux different, and I wanted to understand it. I think
https://lkml.org/lkml/2002/6/3/199 is why I was cc'd on
https://mail.gnome.org/archives/foundation-list/2004-September/msg00014.html but
it's been a while?

Oh, and at this LinuxWorld Expo in 2000:

https://www.fool.com/archive/portfolios/rulemaker/2000/08/18/is-amds-chip-better-than-intels.aspx

I attended a talk where Stallman was VERY UNHAPPY with the "application service
provider" problem where people could run GPL code on a server and not share
modifications because they weren't distributing the code! He already was trying
to figure out how to change the license to stop this, but the problem is since
they weren't distributing but merely using copyright didn't actually give him
any leverage. (He later switched to "tivoization" where he was instead basically
undermining first sale doctrine. That might also be where I learned that GPL
can't rely on contract law for anything because there's no "privity of
contract": you haven't got an agreement if you can't say who your agreement is
WITH. Alas the law has shifted a bit since then in truly horrible ways and the
FSF has been on the wrong side of that...)

So when I launched the first GPL enforcement suits in 2006 it didn't come out of
nowhere. Almost immediately after that the FSF broke the GPL into warring camps
and doomed it to irrelevance, and it took me something like 5 years to work
through the kubler-ross stages of grief. (Maintaining "last GPLv2" versions of
things involved both denial and anger. I did not bargain with the FSF ever
because Rudyard Kipling was right about Danegeld, but there was a lot of
negotation to get SPDX and OSI and github to accept 0BSD...)

> i'd be genuinely curious to know why glibc doesn't have this (and especially
> "why musl does anyway").

Because people were patching a #define __MUSL__ into their toolchains so gnulib
could poke at musl's FILE internals.

https://lists.gnu.org/archive/html/bug-gnulib/2012-06/msg00223.html

Rich has always had a weird reluctance to have a #define admitting you're using
musl. (That's why my mcm-buildall.sh has to append it to features.h after each
toolchain build.) Instead he added __freadahead() so autoconf could probe for
that and use it, avoiding the staircase function entirely.

> i'll have to check what else is in musl's <stdio_ext.h>
> that isn't in glibc's. looks like these:
> 
> size_t __freadahead(FILE *);
> const char *__freadptr(FILE *, size_t *);
> void __freadptrinc(FILE *, size_t);

So far I've just needed to answer the one question...

>       https://linux.die.net/man/3/__fpending
>     <https://linux.die.net/man/3/__fpending>
> 
>     They have _fbufsize() but not how much of the buffer is used. That's just sad.
> 
> that does seem of questionable utility now you bring it up. (though as you
> pointed out on the posix list, unless you _know_ you can seek on a stream, all
> of this is pretty dodgy...)

I need it because I CAN'T seek. If I could, I could just count the bytes
returned by getline() and seek the fd to that position after calling fileno().

The number I'm looking for is "how many bytes can I read from the FILE pointer
without triggering another underlying read from the fd" because I can't re-read
that data from the fd. So I need to fetch the data out of the buffer and handle
it before calling the "takes the rest as a filehandle" function. In the case of
patch.c, something like reading the remaining FILE * data into toybuf() and
doing a write() to the output fd before calling sendfile().

Rob