| 1 |
On Sat, May 30, 2015 at 2:54 PM, Mike Frysinger <vapier@g.o> wrote: |
| 2 |
> we've got a new QA check that warns whenever a package is built using a 32bit |
| 3 |
> filesystem interface. in practice, this applies to arm/mips/ppc/sh/x86 systems |
| 4 |
> (not including multilib -- for now). |
| 5 |
> |
| 6 |
> this topic has come up in Gentoo a few times over the years but we've never |
| 7 |
> really amassed the will power to fix it. instead we fix it in one-offs based |
| 8 |
> on user reports (like "can't download 4GiB file with ftp" #101038). this was |
| 9 |
> worked well enough because most users have moved on to 64bit systems and the |
| 10 |
> interaction with >2GiB files tends to correlate with a few packages. |
| 11 |
> |
| 12 |
> however, "recent" winds have started blowing where file systems are utilizing |
| 13 |
> 64bit inodes to handle large file counts. this means apps that do even basic |
| 14 |
> things like stat() will now fail. the number of applications that this affects |
| 15 |
> is significantly higher, although still relegated to systems that happen to use |
| 16 |
> a file system with 64bit inodes. |
| 17 |
|
| 18 |
Does this issue affect software that only reads/writes small files and |
| 19 |
never calls stat? For example, pkg-config. |
| 20 |
|
| 21 |
It might still be nice to adjust such packages for consistency, but it |
| 22 |
might be harder to justify patches to upstream developers. |
Archives