================
@@ -267,60 +264,47 @@ def update_test(ti: common.TestInfo):
prefix_set = set([prefix for p in run_list for prefix in p[0]])
common.debug("Rewriting FileCheck prefixes:", str(prefix_set))
- for test_id in range(testnum):
- input_line = testlines[test_id]
-
+ for test_id, input_line in enumerate(testlines):
# a {prefix : output, [runid] } dict
# insert output to a prefix-key dict, and do a max sorting
# to select the most-used prefix which share the same output string
p_dict = {}
- for run_id in range(run_list_size):
+ for run_id in range(len(run_list)):
out = raw_output[run_id][test_id]
if hasErr(out):
o = getErrString(out)
else:
o = getOutputString(out)
- prefixes = raw_prefixes[run_id]
-
- for p in prefixes:
+ for p in raw_prefixes[run_id]:
if p not in p_dict:
p_dict[p] = o, [run_id]
- else:
- if p_dict[p] == (None, []):
- continue
+ continue
- prev_o, run_ids = p_dict[p]
- if o == prev_o:
- run_ids.append(run_id)
- p_dict[p] = o, run_ids
- else:
- # conflict, discard
- p_dict[p] = None, []
+ if p_dict[p] == (None, []):
+ continue
- p_dict_sorted = dict(sorted(p_dict.items(), key=lambda item:
-len(item[1][1])))
+ prev_o, run_ids = p_dict[p]
+ if o == prev_o:
+ run_ids.append(run_id)
+ p_dict[p] = o, run_ids
+ else:
+ # conflict, discard
+ p_dict[p] = None, []
# prefix is selected and generated with most shared output lines
# each run_id can only be used once
- used_runid = set()
-
+ used_run_ids = set()
selected_prefixes = set()
- for prefix, tup in p_dict_sorted.items():
- o, run_ids = tup
-
- if len(run_ids) == 0:
- continue
-
- skip = False
- for i in run_ids:
- if i in used_runid:
- skip = True
- else:
- used_runid.add(i)
- if not skip:
+ get_num_runs = lambda item: len(item[1][1])
+ p_dict_sorted = sorted(p_dict.items(), key=get_num_runs, reverse=True)
+ for prefix, (o, run_ids) in p_dict_sorted:
+ if run_ids and used_run_ids.isdisjoint(run_ids):
selected_prefixes.add(prefix)
+ used_run_ids.update(run_ids)
----------------
broxigarchen wrote:
There could be case that the current run_ids list and the used_run_ids is not
completely disjoint. In this case we will not generate the prefix check string,
but still we want to mark these run_id as used.
This is actually an error case. Noted that one run_id can only generate one
output, when duplicated run_id is saw in the prefix iteration, it means there
are other run_id shares the same output, but with different prefix.
An example be like:
```
RUNLINE1 -check-prefix=A
RUNLINE2 -check-prefix=A,B
RUNLINE3 -check-prefix=B
```
runline1,2,3 are the same runline and generate the same output. The dictionary
be like:
```
A: output, [RUNLINE1, RUNLINE2]
B: output, [RUNLINE2, RUNLINE3]
```
RUNLINE2 will be a duplicated run_id. If you generate check-string for both A
and B prefix, the llvm-lit will failed on RUNLINE2, since it will expects two
check-string from RUNLINE2, but one runline only generates one output.
The current code will not generate check-string for B. But I agree we should
throw an error here like what the other update_script does
https://github.com/llvm/llvm-project/pull/164454
_______________________________________________
llvm-branch-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
