1497421 - thread 'main' panicked at 'cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsBaseHashtable.h

Status: RESOLVED INCOMPLETE

(SeaMonkey :: Build Config, defect)

Description

[Edmund Wong (:ewong)]

Getting this for a linux64 trunk build:

38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsISelectionDisplay.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/dom/DOMJSProxyHandler.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/Span.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsJSPrincipals.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsNodeInfoManager.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/malloc_decls.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/mozalloc.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsIArray.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsCSSPropertyID.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/dom/Attr.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nscore.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/dom/TypedArray.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/encoding_rs_statics.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/js/ErrorReport.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/mozilla/TemplateLib.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsIFactory.h
38:27.38 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsDisplayItemTypes.h
40:26.49 thread 'main' panicked at 'cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsBaseHashtable.h
40:26.49 make[3]: *** [force-cargo-library-build] Error 101
40:26.49 make[2]: *** [toolkit/library/rust/target] Error 2
40:26.49 make[1]: *** [compile] Error 2
40:26.49 make: *** [default] Error 2
40:26.49 149 compiler warnings present.
40:26.58 ccache (direct) hit rate: 28.0%; (preprocessed) hit rate: 13.6%; miss rate: 58.4%
40:26.58 /usr/bin/notify-send --app-name=Mozilla Build System Mozilla Build System Build failed

Comment 1

[(not currently active) Ted Mielczarek]

Is that everything you've got in the error output? I would expect there to be an actual backtrace there.

Comment 2

[Edmund Wong (:ewong)]

Attachment: build log

:ted, this is the log.. and as far as I can see, I don't see any
backtraces.  Only warnings.

Also, note of interest is that the rustc package was created locally.
(rationale: I could have just downloaded it from taskcluster, but
TC doesn't have the native OSX64 ones, so I felt doing it locally
would be more helpful in the long run).

Thanks

Comment 3

[(not currently active) Ted Mielczarek]

Yeah, that's not much more informative! I think you do have something wrong with your rust package, FWIW, and that's making the output more confusing.


I think the error is coming from the baldrdash build script, since it's definitely printing an error:
38:02.93      Running `/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/toolkit/library/release/build/baldrdash-23cd65c6995bb497/build-script-build`
38:02.93 cargo:rerun-if-changed=baldrapi.h
38:02.93 cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/js/src/rust/extra-bindgen-flags
38:03.00 Custom { kind: Other, error: StringError("Cannot find binary path") }

https://dxr.mozilla.org/mozilla-central/source/js/src/wasm/cranelift/build.rs

Whatever is failing is failing like this:
40:26.49 thread 'main' panicked at 'cargo:rerun-if-changed=/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/include/nsBaseHashtable.h

...but I think your rust standard library is failing to print the panic message correctly, perhaps because it can't print source info for the panic location? If you use the Rust playground site to run some code like `fn main() { panic!("foo") }` you get the output:
thread 'main' panicked at 'foo', src/main.rs:1:13

I think your Rust stdlib started to print this panic message, but only got up to the first quote and then failed, and then the rest of the line is just output from the stylo build script, which happened to start running after the baldrdash one:
38:03.00      Running `/builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/toolkit/library/release/build/style-00b86e496b84cf9c/build-script-build`

https://dxr.mozilla.org/mozilla-central/source/servo/components/style/build_gecko.rs

The `StringError` there originates from the `which` crate:
https://dxr.mozilla.org/mozilla-esr60/rev/6cdd6d88eca96f6b0725b46c7768997aece49e2b/third_party/rust/which/src/lib.rs#139

..and since the only extern crate that the baldrdash build script uses is bindgen, I'd guess that it's failing here:
https://dxr.mozilla.org/mozilla-esr60/rev/6cdd6d88eca96f6b0725b46c7768997aece49e2b/third_party/rust/bindgen/src/lib.rs#1762

and indeed the only call to `rustfmt_generated_string` in that file will print the error to stderr if it fails:
https://dxr.mozilla.org/mozilla-esr60/rev/6cdd6d88eca96f6b0725b46c7768997aece49e2b/third_party/rust/bindgen/src/lib.rs#1740

...but that doesn't seem to be a fatal error, so I'm still not 100% sure what's happening here.

I did file a cargo issue to make build script output clearer: https://github.com/rust-lang/cargo/issues/6158

Comment 4

[Benjamin Bouvier [:bbouvier] (inactive)]

For what it's worth, this message (Cannot find binary path) seems to appear on many build logs (even on successful builds on treeherder), and I think I've asked on irc in the past and I was told that this message was not meaningful.

Comment 5

[Edmund Wong (:ewong)]

(In reply to Ted Mielczarek [:ted] [:ted.mielczarek] from comment #3)
> Yeah, that's not much more informative! I think you do have something wrong
> with your rust package, FWIW, and that's making the output more confusing.
> 
> 

Thanks for the hints.  I went and redid the packaging but this time I used
1.29.0 instead of 1.29.1.  I'm getting a different error, so I'm guessing
1.29.1 isn't supposed to be used?

0:12.20 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdk.h:33:
 0:12.20 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkcairo.h:51:48: error: unknown type name 'cairo_region_t'
 0:12.20                                          const cairo_region_t *region);
 0:12.20                                                ^
 0:12.25 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkcairo.h:53:1: error: unknown type name 'cairo_region_t'
 0:12.25 cairo_region_t *
 0:12.25 ^
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/xpcom/base/Unified_cpp_xpcom_base2.cpp:47:
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/xpcom/base/nsSystemInfo.cpp:41:
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/system_wrappers/gtk/gtk.h:3:
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gtk/gtk.h:30:
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/system_wrappers/gdk/gdk.h:3:
 0:12.30 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdk.h:46:
 0:12.30 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkpango.h:43:1: error: unknown type name 'cairo_region_t'
 0:12.30 cairo_region_t    *gdk_pango_layout_line_get_clip_region (PangoLayoutLine *line,
 0:12.30 ^
 0:12.35 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkpango.h:48:1: error: unknown type name 'cairo_region_t'
 0:12.35 cairo_region_t    *gdk_pango_layout_get_clip_region      (PangoLayout     *layout,
 0:12.35 ^
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/xpcom/base/Unified_cpp_xpcom_base2.cpp:47:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/xpcom/base/nsSystemInfo.cpp:41:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/system_wrappers/gtk/gtk.h:3:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gtk/gtk.h:30:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/obj-x86_64-pc-linux-gnu/dist/system_wrappers/gdk/gdk.h:3:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdk.h:53:
 0:12.41 In file included from /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdktestutils.h:26:
 0:12.41 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkwindow.h:555:13: error: unknown type name 'cairo_region_t'
 0:12.41                                                 const cairo_region_t *region,
 0:12.41                                                       ^
 0:12.45 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkwindow.h:565:45: error: unknown type name 'cairo_region_t'
 0:12.45                                       const cairo_region_t *shape_region,
 0:12.45                                             ^
 0:12.51 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkwindow.h:592:51: error: unknown type name 'cairo_region_t'
 0:12.51                                             const cairo_region_t *shape_region,
 0:12.51                                                   ^
 0:12.55 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkwindow.h:640:1: error: unknown type name 'cairo_region_t'
 0:12.55 cairo_region_t *gdk_window_get_clip_region  (GdkWindow          *window);
 0:12.55 ^
 0:12.59 /builds/worker/c-cen-t-lnx64-ntly/build/gtk3/usr/local/include/gtk-3.0/gdk/gdkwindow.h:641:1: error: unknown type name 'cairo_region_t'
 0:12.59 cairo_region_t *gdk_window_get_visible_region(GdkWindow         *window);
 0:12.59 ^
 0:12.64 fatal error: too many errors emitted, stopping now [-ferror-limit=]
 0:17.20 20 errors generated.
 0:17.22 make[3]: *** [Unified_cpp_xpcom_base2.o] Error 1
 0:17.22 make[2]: *** [xpcom/base/target] Error 2
 0:17.22 make[2]: *** Waiting for unfinished jobs....

so I'm guessing I'm missing something.

anyway, thanks Ted and Benjamin!  That definitely pointed me in the right direction.  This
unknown type name is a lot easier to work with.

Comment 6

[Frank-Rainer Grahl (:frg)]

Lets close for now and open a new bug if later needed. We are using the standard rust packages on the CentOS and Windows builders and cross compiling macOS.