Don’t ICE if fs::canonicalise fails in meta-load #26044
Conversation
|
r? @nrc (rust_highfive has picked a reviewer for you, use r? to override) |
|
Thanks! Could you add a test for this as well? (probably a |
nagisa
force-pushed
the
canonicalize-metadata-loader
branch
2 times, most recently
from
f5299ba to
5ca401c
Compare
|
Unsure how to make run-make tests run only on specified platforms. |
|
You can take a look at some existing |
This might fail when --extern library is a symlink to an invalid location. Instead just pretend it doesn’t exist at all.
nagisa
force-pushed
the
canonicalize-metadata-loader
branch
from
5ca401c to
ab86fac
Compare
|
Updated. |
| time: libc | ||
| mkdir -p out/time out/time/deps | ||
| ln -sf out/libc/liblibc.rlib out/time/deps/ | ||
| $(RUSTC) in/time/lib.rs -Ldependency=out/time/deps/ |
There was a problem hiding this comment.
Isn't this command expected to fail without an ICE error message? (e.g, shouldn't there be a piping or assertion of some form here?)
There was a problem hiding this comment.
As far as I understand -Ldependency= adds some path to a list of directories in which to look for library files, right? Then this fails to use symlinked libc and proceeds to find one in rustc’s own distribution, which is why this doesn’t fail.
If this specified --extern libc=, though, then it would be a different problem.
There was a problem hiding this comment.
Ah right, I see what's going on here now, nevermind then!
|
⌛ Testing commit ab86fac with merge 295ef21... |
|
💔 Test failed - auto-mac-64-opt |
|
@bors: retry On Mon, Jun 8, 2015 at 12:35 AM, bors [email protected] wrote:
|
…chton This might fail when --extern library is a symlink to an invalid location. Instead just pretend it doesn’t exist at all. Fixes #26006
This might fail when --extern library is a symlink to an invalid location. Instead just pretend it
doesn’t exist at all.
Fixes #26006