| 1 |
On Thu, Aug 18, 2016 at 1:39 AM, Daniel Campbell <zlg@g.o> wrote: |
| 2 |
> |
| 3 |
> Is it as simple as switching the linker and re-merging packages that one |
| 4 |
> maintains? Is gold supposed to be a big deal? Does it do the job of |
| 5 |
> linking better? I read the blog post and all but nobody's explaining |
| 6 |
> what gold does better than standard ld. |
| 7 |
> |
| 8 |
|
| 9 |
There are a bunch of reasons why people prefer it for the long term. |
| 10 |
However, one of the biggest drives to get devs to use it is that it |
| 11 |
serves as a good canary for underlinking. |
| 12 |
|
| 13 |
If you underlink with the bfd linker the package will build and run |
| 14 |
fine typically, at least initially. However, down the road as |
| 15 |
libraries are updated you can run into problems. Since you didn't |
| 16 |
specify all your dependencies portage won't do rebuilds when they're |
| 17 |
necessary. My understanding is that if the needed libraries aren't |
| 18 |
even linked then even preserve-rebuild will fail. The package is |
| 19 |
basically relying on unspecified behavior to work. If you need a |
| 20 |
symbol you're supposed to link to the providing library, not just |
| 21 |
assume that it will happen to be around because some other library |
| 22 |
you're linked to happens to pull it in. |
| 23 |
|
| 24 |
The gold linker generates errors if you attempt to underlink, which |
| 25 |
turns this into a build-time error, and not a |
| 26 |
maybe-you-see-it-maybe-you-don't runtime issue for random users in six |
| 27 |
months. |
| 28 |
|
| 29 |
Correctly specifying dependencies and ensuring they're linked is of |
| 30 |
benefit to all users, and it will prevent subtle problems for users of |
| 31 |
the default bfd linker. |
| 32 |
|
| 33 |
At least, that is my understanding of the issue. It is entirely |
| 34 |
possible I missed something; I don't profess to be an ELF expert. :) |
| 35 |
|
| 36 |
-- |
| 37 |
Rich |
Archives