This information is strictly available in the log (you can find the original
call) but it makes the log easier to follow in practice.
Details
Details
Diff Detail
Diff Detail
- Repository
- rL LLVM
Event Timeline
Comment Actions
Log now looks something like this:
I[16:47:01.225] <-- textDocument/completion("2") I[16:47:01.233] Ignored diagnostic. /usr/local/google/home/sammccall/temp/test.cc:7:1:'main' must return 'int' I[16:47:01.234] Code complete: sema context Statement, query scopes [] (AnyScope=false) I[16:47:01.234] Code complete: 0 results from Sema, 0 from Index, 0 matched, 0 returned. I[16:47:01.234] --> reply:textDocument/completion("2") 8 ms
Comment Actions
LG, thanks!
clangd/ClangdLSPServer.cpp | ||
---|---|---|
218 ↗ | (On Diff #170890) | Move locking below to not lock while copying reply. |
221 ↗ | (On Diff #170890) | What about also printing vlog after sending out reply? Because for "heavy" replies there is a huge difference between process time and serialization of response to json, sometimes it might help to see how much time is spent for serialization as well.(for example ycmd patch, it takes a few hundred ms to prepare reply, but then a few seconds to dump it) |
clangd/ClangdLSPServer.cpp | ||
---|---|---|
221 ↗ | (On Diff #170890) | As discussed offline, this is interesting, but only relevant for really huge responses and probably noise the rest of the time. |