| Summary: | Stap does not check incorrect probe point syntax | ||
|---|---|---|---|
| Product: | systemtap | Reporter: | Masami Hiramatsu <mhiramat> |
| Component: | translator | Assignee: | Unassigned <systemtap> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | ||
| Priority: | P2 | ||
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Host: | Target: | ||
| Build: | Last reconfirmed: | ||
| Attachments: | A bugfix patch for wildcard parsing. | ||
Created attachment 2111 [details]
A bugfix patch for wildcard parsing.
This patch will fix the first and the second issue.
With this patch, parser treats '*'(wildcard) in probe points as an alphabet.
However, there is a side-effect that this patch allows consecutive wildcards
(ex. "probe a** {}").
I'm not sure why the consecutive wildcards have been restricted. I think
it could be allowed like as bash.
commit 1ad820e finally fixes issue 3. |
- 'probe b* g* n {}' is recognized as 'b*g*n' - 'probe hoge* if(e) {}' is recognized as 'hoge*if()' (concatenate 'hoge*' and 'if') - 'probe huga? = hoge {}' is recognized as 'huga = hoge' (? is just ignored) I think all those cases should be treated as incorrect syntaxes. Thanks,