diff options
| author | Peter Klausler <pklausler@nvidia.com> | 2022-06-10 18:08:14 -0700 |
|---|---|---|
| committer | Peter Klausler <pklausler@nvidia.com> | 2022-06-15 17:29:50 -0700 |
| commit | cfbde7149d8647a4ff2eef892a30ffc55605df2c (patch) | |
| tree | c61e0664cfeca69d20492c19c20fa78e4d4e7cbb /lldb/test/Shell/ScriptInterpreter/Python | |
| parent | 99796d06dbe11c8f81376ad1d42e7f17d2eff6ae (diff) | |
| download | llvm-cfbde7149d8647a4ff2eef892a30ffc55605df2c.zip llvm-cfbde7149d8647a4ff2eef892a30ffc55605df2c.tar.gz llvm-cfbde7149d8647a4ff2eef892a30ffc55605df2c.tar.bz2 | |
[flang][runtime] Catch more (all?) negative unit number errors
Fortran does have negative unit numbers -- they show up in child I/O
subroutines for defined I/O and for OPEN(NEWUNIT=) -- but the runtime
needs to catch the cases where a negative unit number that wasn't
generated by the runtime is passed in for OPEN or for an I/O statement
that would ordinarily create an anonymous "fort.NNN" file for a
hitherto unseen unit.
Differential Revision: https://reviews.llvm.org/D127788
Diffstat (limited to 'lldb/test/Shell/ScriptInterpreter/Python')
0 files changed, 0 insertions, 0 deletions