Commit 76b6266
authored
Fix the diagnostic verifier, plus a few other fixes. (#598)
* Reverse name of 3c -disccty option to reflect what it actually does.
By default, 3c adds the -f3c-tool compiler option to disable the Checked
C type checker. 3c's -disccty option (apparently an abbreviation for
"disable Checked C type checker") made 3c _not_ pass -f3c-tool, hence
leaving the type checker _enabled_. 3C's internal flag variable,
DisableCCTypeChecker, was also backwards, though interestingly, the
`3c -help` text was correct ("Do not disable checked c type checker").
This commit renames -disccty to -enccty and renames DisableCCTypeChecker
to EnableCCTypeChecker. We take the opportunity to further improve the
`3c -help` text too.
* Completely remove the old `3c -verify` option.
Now that 3C only runs one compiler "invocation" per translation unit,
with the new diagnostic verifier implementation, callers will be able to
just use the `-Xclang -verify` compiler option. And given that we may
want to use other diagnostic verifier options (e.g., `-verify=PREFIX` or
`-verify-ignore-unexpected`), let's standardize on using the original
compiler options rather than trying to define `3c` options wrapping all
of them.
* Switch from ClangTool::buildASTs to a custom _3CASTBuilderAction.
Replace the existing ArgumentsAdjusters with modifications to the option
data structures in _3CASTBuilderAction. More will be added to
_3CASTBuilderAction in the subsequent commits.
Fixes #536.
* Canonicalize -I directory paths in _3CASTBuilderAction.
This works per translation unit, so hopefully it is a complete and
correct fix to #515 that benefits all callers of 3C.
Remove the previous crude workaround from convert_project.
Fixes #515.
* Make diagnostic verification work after the redesign of #488.
Re-enable diagnostic verification in regression tests where it still
succeeds. It remains disabled in some tests in which failures were
introduced while it was disabled; #609 is for that.
Fixes #503.
* Migrate the difftypes* tests to the diagnostic verifier.
This feels a little cleaner than manually scanning the stderr and lets
us check the diagnostics more precisely, for what that's worth. We
couldn't do this before #488 both because 3C exited before the
diagnostics could be verified and because 3C didn't support custom
verify prefixes. #488 solved the first problem, and this PR solves the
second.
* Document diagnostic verification in 3C's CONTRIBUTING.md.1 parent e549c91 commit 76b6266
File tree
19 files changed
+432
-247
lines changed- clang
- docs/checkedc/3C
- include/clang/3C
- lib/3C
- test/3C
- base_subdir
- tools/3c
- utils/port_tools
19 files changed
+432
-247
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
94 | 94 | | |
95 | 95 | | |
96 | 96 | | |
| 97 | + | |
| 98 | + | |
| 99 | + | |
| 100 | + | |
| 101 | + | |
| 102 | + | |
| 103 | + | |
| 104 | + | |
| 105 | + | |
| 106 | + | |
| 107 | + | |
| 108 | + | |
| 109 | + | |
| 110 | + | |
| 111 | + | |
| 112 | + | |
| 113 | + | |
| 114 | + | |
| 115 | + | |
| 116 | + | |
| 117 | + | |
| 118 | + | |
| 119 | + | |
| 120 | + | |
| 121 | + | |
| 122 | + | |
| 123 | + | |
| 124 | + | |
| 125 | + | |
| 126 | + | |
| 127 | + | |
| 128 | + | |
| 129 | + | |
| 130 | + | |
| 131 | + | |
| 132 | + | |
| 133 | + | |
| 134 | + | |
| 135 | + | |
| 136 | + | |
| 137 | + | |
| 138 | + | |
| 139 | + | |
| 140 | + | |
| 141 | + | |
| 142 | + | |
| 143 | + | |
| 144 | + | |
| 145 | + | |
| 146 | + | |
| 147 | + | |
| 148 | + | |
97 | 149 | | |
98 | 150 | | |
99 | 151 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
57 | 57 | | |
58 | 58 | | |
59 | 59 | | |
60 | | - | |
| 60 | + | |
61 | 61 | | |
62 | 62 | | |
63 | 63 | | |
| |||
68 | 68 | | |
69 | 69 | | |
70 | 70 | | |
71 | | - | |
72 | | - | |
73 | | - | |
74 | | - | |
75 | | - | |
76 | 71 | | |
77 | 72 | | |
78 | 73 | | |
| |||
104 | 99 | | |
105 | 100 | | |
106 | 101 | | |
| 102 | + | |
| 103 | + | |
107 | 104 | | |
108 | 105 | | |
109 | 106 | | |
| |||
141 | 138 | | |
142 | 139 | | |
143 | 140 | | |
| 141 | + | |
| 142 | + | |
| 143 | + | |
| 144 | + | |
| 145 | + | |
144 | 146 | | |
145 | 147 | | |
146 | 148 | | |
147 | | - | |
| 149 | + | |
| 150 | + | |
| 151 | + | |
| 152 | + | |
| 153 | + | |
| 154 | + | |
| 155 | + | |
| 156 | + | |
| 157 | + | |
| 158 | + | |
| 159 | + | |
148 | 160 | | |
149 | 161 | | |
150 | 162 | | |
| |||
0 commit comments