I have used ClearCase for 20 years and this is the a crazy problem.
I have encountered it with another compiler 8 years ago but this time I have no work-around.
FileA includes HeaderX and HeaderY but the precompiler never reads HeaderY which has nothing to do with HeaderX
(take HeaderX out of the equation or make relatively arbitrary changes to HeaderX and HeaderY is read)
Randomly, we are seeing the old powerpc-wrs-vxworks gcc-2.96 compiler not include files during precompilation.
This only happens in a particular vob and only happens in dynamic views and occurs randomly.
We believe that this is specific to this particular VOB and believe that there is something wrong with the VOB.
We have experienced the problem with Win7 ClearCase version 7.1.2.1, but also with WinXP.
Over the years, we have noticed a similar problem with this VOB with another compiler.
When we traced the compiler, we see that it opens and fstat's the include file and then closes without a read.
Normally, the compiler will perform a read after the open. The trace shows the fstat returning a valid size.
Below is the simplest distillation of the problem but do not be lured into thinking it is something about the
contents of the file for an explanation of the behavior - you can make arbitrary changes to the files to elicit
the correct behavior. Specifically, you can use identical files moved to a different location in the same view or outside
the view and the compiler will behave correctly or use a different view with the same files and it will work.
We believe, some interaction with MVFS, specific to this VOB, and the compiler is the problem.
Has anyone seen a state that a malfunctioning VOB could get into that may exhibit this misbehavior randomly in views?
foo.c:
#include "HeaderX.h" /* fix 1 remove this line
fix 2 rename this include file
fix 3 move this include file to a new dir */
#include "HeaderY.h" /* fix 4 move this line up
fix 5 rename this include file
fix 6 move this include file to a new dir */
HeaderX.h:
/* a=b; fix 7 uncomment this line */
#ifndef X /* fix 8 change this to an #ifdef instead of #ifndif */
#define X /* fix 9 remove this line
fix 10 change just this X to Y
fix 11 move define after endif */
#endif
HeaderY.h:
#error hurrah! I was included /* GOAL -- to have this file included
If it is included the precompiler
will print this message.
*/
P.S Don't tell me it is a compiler bug.
Reasons to believe it is a ClearCase issue:
Only happens in dynamic views
Happens on multiple machines, multiple users, multiple views, different operating systems. But it is somewhat random.
Reasons to doubt my reasons:
Fixes 7, 8, 9, 10, 11
In cases where it fails, using a different version of the compiler will not cause the error.
But I think questioning the compiler is a red herring, the same compiler with the same set of files will work fine outside of ClearCase or in a different directory in the same view. The VOB is about 15 years old and I think it is peculiar to the VOB. But it is a huge effort to replace it.
Without replacing the Vob, is it possible to export a subset of that Vob representing the sources of that specific project in a brand new Vob?
A simple clearfsimport would create a fresh history, while the original Vob would still be used for:
history consultation for those files exported in the new Vob
everything else as usual for all the other files not exported in the new Vob.
I have see a few case of Vob corruption, and that option was the easiest workaround.
Related
I found a pacman project in github where a file conf.c includes a header file #include "ini.h" where ini.h contains only a single line (i.e no #include statement):
//ini.h
../common/ini.c
I have never seen anyone doing this before! It seems a bit hackish/rough around the edges. My questions are:
Is this legal C?
Is it portable?
Is it recommended?
I would have assumed the answer should be no for all these questions, but I may be learning something new...
edit
From the answers, I see its a Linux symlink. I guess that this means it is not portable to Windows, and would also make it more difficult to read outside a unix environment. I would also imagine that using relative paths (or include directories) instead of symlinks would be a better practice in cases like this for reasons mentioned above...
src/pacman/ini.h is a symbolic link according to the site.
Symbolic link has an information of where the target file is (path name), and I guess it is what is displayed on the site.
The OS will redirect access to that ini.h to ../common/ini.h, which is a normal C code.
I don't see any reason why not. The include statement indicates the compiler to replace that line with the whatever is in the included file
I have to work with a C/C++ build environment that drops intermediate files all over the place:
.i files containing the output of the C-preprocessor (roughly raw C)
.s files containing the input of the C-assembler
CEDET (I assume the semantic analyzer) eventually finds these files and attempts to index them. This results in jumping to .i files containing raw C for definitions and generally slowing down parsing and loading of the .semanticdb.
I never open these files in emacs, so they must be being loaded by the background analyser.
Is it possible to prevent the analyser from loading these files? I can't find any configuration options that define the file-types that are parsed by the background analyser.
If you never need C mode for these files, here's a quick fix:
(add-to-list 'auto-mode-alist '("\\.i\\'" . fundamental-mode))
(add-to-list 'auto-mode-alist '("\\.s\\'" . fundamental-mode))
The answer from abo-abo gave me the clues I needed. The grep implementation (used by EDE) of semantic-symref-perform-search uses auto-mode-alist to find matching files for a given semantic mode (based on the current buffer's mode - eg `c-mode) when trying to resolve symbols.
The final fix I used is to specifically eliminate the default entries in the auto-mode-alist using:
(delete '("\\.i\\'" . c-mode) auto-mode-alist)
(delete '("\\.ii\\'" . c++-mode) auto-mode-alist)
Adding fundamental-mode entries as suggested by abo-abo seems to work also, however I was concerned that since the c-mode entries were still in the list a change in implementation could result in them being reactivated.
On OS X, I generated a set of ctags for the system includes using the following command:
ctags -f c -h ".h" -R --c-kinds=+p --fields=+iaS --extra=+q /usr/include
This was run inside of a ~/.vim/ctags/ directory, where I put all of the ctags I generate for system-wide header files (I also have stuff for ROS and CPP that I load conditionally, but that's neither here nor there).
Anyway. The ctags file is set correctly in my .vimrc, and vim can definitely see the ctags, but for some reason the autocomplete popup will only display results from #included header files if I write out the entire symbol and then start backspacing. As an example, if I #include <string.h> in a project, and then I want to call strlen(), and I start to type str in to the active vim buffer, I will only get results for symbols that are currently in the vim buffer. But, if I type out strlen and then start backspacing one or two characters and hit <C-n>, the popup menu will be populated with matches from any other included header files.
EDIT: Turns out, if I just hit "s" then <C-n>, it works as well. So the problem seems to be that it only works if the popup menu is launched manually. Which makes me think that it's a plugin problem (see below)
Additional information:
completeopt is set to completeopt=menuone,menu,preview,longest
I have OmniCppComplete, which I suppose could be interfering with the behavior. It is currently not being conditionally loaded for C++ files only. If you want me to edit and post my OmniCppComplete settings from my .vimrc, just ask.
I also have AutoComplPop installed, but I haven't done anything to configure it, so it's running with its default settings. Haven't really researched the plugin, so no idea if some of it's behavior could be interfering with the results.
I have AutoTag and TagBar installed, but those should only be fiddling with the current directory's local tagfile.
I'm honestly pretty new to Vim, and I just have no idea where to start debugging this issue, whether it be with a random plugin or with my .vimrc settings.
Vim has many specific completion mechanisms.
<C-n> and <C-p> use many sources defined by the complete option. By default, they will provide completion using the current and all loaded and unloaded buffers, tags and included files. While you can usually get quite useful suggestions with these, it is a bit of a "catch-all" solution: it is not reliable at all if you work on reasonably large projects.
<C-x><C-]> uses only tags so it may be a little more useful to you.
And there are many more, see :h ins-completion.
Omni completion is smarter: it typically runs a custom filetype-specific script that tries hard to provide meaningful completion. It is triggered by <C-x><C-o> and you can read about it in :h ft-c-omni. Omni completion is often a better choice when working with code.
Because you have two overlaping "auto"-completion plugins it's hard to say what completion mechanism is at work. You should disable those plugins and play around with the different completion mechanisms available to you.
I have not mastered this yet, but I do think the following observation may be of help.
Vim's default auto complete which can be quite noisy, often gets in the way of what you call with <C-x><C-o>. Specifically, I found myself calling up my tags based completions with <C-x><C-o> only to have them replaced with continued typing with Vim's default suggestions using my open buffers.
The suggestion of shutting off one of the plugins makes sense. In my case the key was how to shut down Vim's default behavior. I have seen several people (and to which I now include myself), set the length of the expression to a high number before triggering Vim's default. For me that is:
let g:deoplete#auto_complete_start_length = 99
... this way you eliminate the default layer of completions that comes and goes regardless of the commands you intended to inform your work.
This still feels like a hack but it helps keep my work focused on the tag-based completions.
FYI: I use NVIM on a Mac.
I'm developing something on Minix 3 and, when it comes to deal with io files, I got a problem.
In the code:
#include <stdio.h> /* If I don't call any stdio funcs compiler doesnt's complain*/
int main() {
FILE * fp; /* I get the following: " * not expected " */
return 0;
}
Already tried everything that comes to my mind, can't figure it out..
/EDIT/
From what I can tell, when I include something, if I call functions not related to structs, it's OK. Is it the structs ?
I will assume you have checked whether the Minix file is present, that it really defines the type FILE and that your include path provides the correct -Ioption to the compiler to find that file.
Depending on your environment it could happen that an environment variable INCLUDE exists and is recognized by your compiler to provide additional include paths, recognized even before the include options from the command line. In such a case it might happen to include a stdio.hfrom a different compiler. Visual Studio is known to provide such an environment variable by default, and that has bitten me once before.
EDIT: Running the preprocessor in isolation may help to find out what is really happening in any case. Verify that FILEis defined in the preprocessed version of your file.
I have a huge project, whole written in C language and I have a single make file that is used to compile it. The project C files contains lots of capitalize problems in it's header files, meaning there are tones of header files that were miss-spelled in lots of C files.
The problem is I need to migrate this project to compile on Linux machine and since Linux is case sensitive I got tones of errors.
Is there an elegant way which I can run make file in Linux and tell him to ignore case sensitive?
Any other solution will be welcome as well.
Thanks a lot.
Motti.
You'll have to fix everything by hand and rename every file or fix every place with #include. Even if you have a huge project (comparable with linux kernel), it should be possible to do this during a hour or two. Automation may be possible, but manual way should be better - because script won't be able to guess which name is right - filename, or the name used in #include.
Besides, this situation is a fault of original project developer. If he/she wasn't sloppy and named every header in every #include correctly, this wouldn't happen. Technically, this is a code problem similar to syntax error. The only right way to deal with it is to fix it.
I think it takes not too long to write a small script, which goes thru the directories first, then replaces C headers. Explained:
Scan the headers' folder and collect filenames.
Make a lowercase list of them. You have now original and locased pairs.
Scan the C source files and find each line contains "#include"
Lowercase it.
Find the lowercase filename in the list collected and lowercased from headers.
Replace the source line with the one collected from headers.
You should put the modified files into a separate folder structure, avoid overwriting the whole source with some buggy stuff. Don't forget to create target folders during the source tree scan.
I recommend a script language for that task, I prefer PHP, but just it's the only server-side script language which I know. Yep, it will run for a while, but only once.
(I bet that you will have other difficulties with that project, this problem is not a typical indicator of high quality work.)
Well I can only tell you that you need to change the case of those header files. I don't know that there is any way you can make it automatic but still you can use cscope to do it in a easier way.
http://www.linux-tutorial.info/modules.php?name=ManPage&sec=1&manpage=cscope
You can mount the files on a case-insensitive file system. FAT comes to mind. ntfs-3g does not appear to support this.
I use the find all and replace all functionality of Source Insight when i have to do complete replacement. But your problem seems quite big, but you can try the option to replace every header file name in all occurences of source files using the
"Find All" + "Replace" functionality. You can use notepad++ too for doing the same.
A long time ago there was a great tool under MPW (Macintosh Programmer's Workshop) called Canon. It was used to canonize text files, i.e. make all symbols found in a given refernce list have have the same usage of upper/lower case. This tool would be ideal for a task like this - I wonder if anything similar exists under Linux ?