#38018 closed defect (fixed)
Doxygen: Prevent clang memory leak by adding back the flag reinplaced away by r101062
| Reported by: | cooljeanius (Eric Gallager) | Owned by: | cssdev |
|---|---|---|---|
| Priority: | Normal | Milestone: | |
| Component: | ports | Version: | 2.1.3 |
| Keywords: | Cc: | mamoll (Mark Moll), raimue (Rainer Müller) | |
| Port: | doxygen |
Description (last modified by larryv (Lawrence Velázquez))
Removing -Wno-invalid-source-encoding causes clang++ to memory leak when it hits certain files used when compiling doxygen, which makes it take forever, to the point where I usually just end up ^C-ing it. I can see why removing this flag might be necessary for gcc, but couldn't this reinplace at least be done conditionally? i.e. do the reinplace iff configure.compiler == gcc; on the other hand, if configure.compiler == clang, then don't do it.
(r101062 for reference)
Change History (12)
comment:1 Changed 12 years ago by larryv (Lawrence Velázquez)
| Cc: | mmoll@… removed |
|---|---|
| Description: | modified (diff) |
| Owner: | changed from macports-tickets@… to mmoll@… |
comment:2 Changed 12 years ago by cooljeanius (Eric Gallager)
Since my main.log got overwritten once I finally properly fixed my local copy of my portfile, here's some diagnostic information btw:
gl00b05045:~ root# uname -amnprsv Darwin gl00b05045 11.4.2 Darwin Kernel Version 11.4.2: Thu Aug 23 16:25:48 PDT 2012; root:xnu-1699.32.7~1/RELEASE_X86_64 x86_64 i386 gl00b05045:~ root# which clang /usr/bin/clang gl00b05045:~ root# clang --version Apple LLVM version 4.2 (clang-425.0.24) (based on LLVM 3.2svn) Target: x86_64-apple-darwin11.4.2 Thread model: posix gl00b05045:~ root# xcodebuild -version Xcode 4.6 Build version 4H127
comment:3 follow-up: 4 Changed 12 years ago by mamoll (Mark Moll)
It's easy enough to fix with something like this:
if {[string match *gcc* ${configure.compiler}]} {
...
}
but -Wno* flags only suppress warnings, right? They shouldn't affect code generation, I thought. Are you sure this is the problem? Also, why is this assigned to me? Isn't css the maintainer for this port?
comment:4 Changed 12 years ago by larryv (Lawrence Velázquez)
| Cc: | mmoll@… added; css@… removed |
|---|---|
| Owner: | changed from mmoll@… to css@… |
Replying to mmoll@…:
Also, why is this assigned to me? Isn't css the maintainer for this port?
Sorry, I figured you might know something about it, since you made the -Wno-invalid-source-encoding change.
comment:5 Changed 12 years ago by mamoll (Mark Moll)
Ah, I forgot about that. I can reproduce the bug. Fixed in r103015.
comment:6 Changed 12 years ago by cooljeanius (Eric Gallager)
Thanks! (This can be closed as "fixed" now btw)
comment:7 Changed 12 years ago by raimue (Rainer Müller)
| Cc: | raimue@… added |
|---|
Has this been reported upstream for clang++? Can we reproduce it using a minimal example (at best using just one source file)?
comment:8 Changed 12 years ago by mamoll (Mark Moll)
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
I don't have time to track down this bug in clang++. Sorry, closing this issue.
comment:9 Changed 12 years ago by jmroot (Joshua Root)
Someone please at least file a radar. Someone reported the same thing on the Fink ML and said the problem was introduced in Xcode 4.6.
comment:10 Changed 12 years ago by mamoll (Mark Moll)
Ok, bug filed: https://bugreport.apple.com/cgi-bin/WebObjects/RadarWeb.woa/54/wo/AWrUtraQlHBqiEGKvYiVI0/19.66 (Not sure if this URL will work, but the Problem ID is 13202404.)
Thanks.