diff options
author | Sam Clegg <sbc@chromium.org> | 2017-06-21 18:05:57 -0700 |
---|---|---|
committer | GitHub <noreply@github.com> | 2017-06-21 18:05:57 -0700 |
commit | c0ae2e69b53f12e57833270e1b48a01864fb5156 (patch) | |
tree | f296a04b560602ab1a02877c01ce617d4d48dd3f /src/binary-reader-interpreter.cc | |
parent | 9c7b8e5cfe4f7c2dda32f21167231d2653da2c7c (diff) | |
download | wabt-c0ae2e69b53f12e57833270e1b48a01864fb5156.tar.gz wabt-c0ae2e69b53f12e57833270e1b48a01864fb5156.tar.bz2 wabt-c0ae2e69b53f12e57833270e1b48a01864fb5156.zip |
Allow duplicate names in the name section (#512)
While this was previously allowed, it would result
in an invalid wast file since mutliple functions
would end up with the same name.
When generating labels we now add a unique numeric
suffix (.1, .2 etc) a la llvm-link.
We could add magic behaviour to strip these out
again when producing the name sectio in wast2wasm
which would allow for cleaner roundtrip, but might
cause confusion/issues for those who might want
to legitimately name their functions with such
suffixes.
Diffstat (limited to 'src/binary-reader-interpreter.cc')
0 files changed, 0 insertions, 0 deletions