| 1 |
On 31 May 2015 11:58, Mike Gilbert wrote: |
| 2 |
> On Sat, May 30, 2015 at 2:54 PM, Mike Frysinger wrote: |
| 3 |
> > we've got a new QA check that warns whenever a package is built using a 32bit |
| 4 |
> > filesystem interface. in practice, this applies to arm/mips/ppc/sh/x86 systems |
| 5 |
> > (not including multilib -- for now). |
| 6 |
> > |
| 7 |
> > this topic has come up in Gentoo a few times over the years but we've never |
| 8 |
> > really amassed the will power to fix it. instead we fix it in one-offs based |
| 9 |
> > on user reports (like "can't download 4GiB file with ftp" #101038). this was |
| 10 |
> > worked well enough because most users have moved on to 64bit systems and the |
| 11 |
> > interaction with >2GiB files tends to correlate with a few packages. |
| 12 |
> > |
| 13 |
> > however, "recent" winds have started blowing where file systems are utilizing |
| 14 |
> > 64bit inodes to handle large file counts. this means apps that do even basic |
| 15 |
> > things like stat() will now fail. the number of applications that this affects |
| 16 |
> > is significantly higher, although still relegated to systems that happen to use |
| 17 |
> > a file system with 64bit inodes. |
| 18 |
> |
| 19 |
> Does this issue affect software that only reads/writes small files and |
| 20 |
> never calls stat? For example, pkg-config. |
| 21 |
|
| 22 |
it is possible to use some file system operations and never run into problems. |
| 23 |
the trouble is that it requires diligence and careful auditing, and you have to |
| 24 |
make sure your interactions with other packages are also OK. considering how |
| 25 |
bad/unaware the wider community is wrt LFS, ABI compatibility, and such critical |
| 26 |
but oft overlooked issues, i suspect most people also won't have the will |
| 27 |
power/capabilities to keep up here. |
| 28 |
|
| 29 |
conversely, flip the autoconf/cppflags magic, and most projects Just Work. |
| 30 |
|
| 31 |
> It might still be nice to adjust such packages for consistency, but it |
| 32 |
> might be harder to justify patches to upstream developers. |
| 33 |
|
| 34 |
pkg-config already merged it and it's already been pulled back into Gentoo ;) |
| 35 |
https://bugs.freedesktop.org/show_bug.cgi?id=90078 |
| 36 |
https://bugs.gentoo.org/550508 |
| 37 |
|
| 38 |
that said, i'm not entirely sure why there would be a lot of push back. |
| 39 |
when i gathered data on this years ago (which i no longer have, so i'd have |
| 40 |
to recreate), the increased foot print and runtime cost was not significant. |
| 41 |
especially when you consider that we're talking about I/O operations which |
| 42 |
inherently kill CPU performance due to stalls. |
| 43 |
-mike |
Archives