Alas, you missed a serious bug in that same function. Patch attached to fix it and improve some comments. No more late night coding for me :) That leaves Path.cpp. Once that's done I can actually try and execute this stuff. On Thu, 16 Sep 2004 08:55:28 -0700 Reid Spencer <reid at x10sys.com> wrote:> This version has been committed to CVS with one minor change: > I made the warning about caller calling LeaveCriticalSection a little > more prominent. Patch is here: > > http://mail.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20040913/018371.html > > Reid.-------------- next part -------------- A non-text attachment was scrubbed... Name: Signals.diff Type: application/octet-stream Size: 1572 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20040916/62975626/attachment.obj>
Jeff, Your patch has been committed. Thanks! Reid. Jeff Cohen wrote:> Alas, you missed a serious bug in that same function. Patch attached to > fix it and improve some comments. No more late night coding for me :) > > That leaves Path.cpp. Once that's done I can actually try and execute > this stuff. > > > On Thu, 16 Sep 2004 08:55:28 -0700 > Reid Spencer <reid at x10sys.com> wrote: > > >>This version has been committed to CVS with one minor change: >>I made the warning about caller calling LeaveCriticalSection a little >>more prominent. Patch is here: >> >>http://mail.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20040913/018371.html >> >>Reid. >> >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>LLVM Developers mailing list >>LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu >>http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev
Here's my first take on Path.cpp. It has a few holes still but it ought to be basically functional. A few observations: First, the Win32 version still has a lot of code in common with Unix. This code should be pulled out into a common file. Second, some of the functions leave me scratching me head. What purpose does create_file serve? Is there really a need to create empty files? (It also leaks an open file handle, BTW). I would expect set_file to alter only the file name portion of the path and leave the directory portion unchanged, and the opposite for set_directory. set_directory obliterates the file name portion of the path (OK, I guess) but set_file doesn't make any sense at all. -------------- next part -------------- A non-text attachment was scrubbed... Name: Path.diff Type: application/octet-stream Size: 10799 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20040918/681287f5/attachment.obj>