I have a particular executable (let's call it bin) which crashes with a segfault when run normally with ./bin, but if I create this Makefile:
all:
./bin
and I do make, the executable runs without error and terminates correctly. How is this even possible?
You don't say what version of make you're using, but older versions of GNU make had a bug where make set its own stack size to "unlimited" then didn't set it back to the default value when running programs. Newer versions of GNU make fix that bug so that programs are run with the default stack size.
See https://savannah.gnu.org/bugs/?func=detailitem&item_id=22010
Related
I'm learning GNU make . Suppose I have a hello_world.c file and a Makefile:
hello_world.c:
#include <stdio.h>
int main() {
printf("Hello World!\n");
return 0;
}
Makefile:
hello: hello_world.c
gcc hello_world.c -o hello_world
Now, I think hello is my target and hello_world.c is its dependency. If make somehow detects that hello_world.c is newer than its object-file, it executes the corresponding command.
1- How does make manage file version control and how does it detect something is newer than something else and needs updating?
2- If I change hello_world.o using an editor and corrupt the file, it obviously does not execute but make hello reports that nothing needs to be done!
I mean, make only checked that the dependency is older than the target and exited doing nothing. I think it should have detected that this target is not the one corresponding to its latest invocation. Somehow, it should have compared the "combination of dependency AND target" history instead of just comparing dependency's history w.r.t. target.
3- Is this the limitation of make? How can I circumvent this issue? Because there maybe some external apps messing with the target of my make operation.
1- How does make manage file version control and how does it detect something is newer than something else and needs updating?
You already know the answer to this question. You said it in your next paragraph: "I mean, make only checked that the dependency is older than the target and exited doing nothing." That's right. make updates dependent targets when the dependencies of targets are newer.
2- If I change hello_world.o using an editor and corrupt the file, it obviously does not execute but make hello reports that nothing needs to be done! I mean, make only checked that the dependency is older than the target and exited doing nothing. I think it should have detected that this target is not the one corresponding to its latest invocation. Somehow, it should have compared the "combination of dependency AND target" history instead of just comparing dependency's history w.r.t. target.
You are asking make to do a lot more than what it was intended to do.
3- Is this the limitation of make? How can I circumvent this issue? Because there maybe some external apps messing with the target of my make operation.
From your point of view, it does seem like it is a limitation of make. However, I want to point out that you are sabotaging the workings of make by updating a target manually.
How can you circumvent it?
Don't manually modify targets that are built by make.
Manually update the time stamp of one of the dependencies and then run make. You can use the command touch for that.
Provide a dummy target named clean that will remove all the dependent targets. Then, run make clean followed by make.
Provide a dummy target named rebuild. Forcefully build whatever you need to build in that target. Then, run make rebuild.
The make program simply compare the modification time-stamps of the target (hello) and the dependency (hello_world.c) files.
If the time-stamp of a dependency file is newer than the targets, then it execute the commands.
How does make manage file version control and how does it detect something is newer than something else and needs updating?
Very simple: make does not care about file versions.
Detecting if something is newer than something else is simply done by comparing time stamps from file system.
If I change hello_world.o using an editor and corrupt the file, it obviously does not execute but make hello reports that nothing needs to be done! I mean, make only checked that the dependency is older than the target and exited doing nothing. I think it should have detected that this target is not the one corresponding to its latest invocation. Somehow, it should have compared the "combination of dependency AND target" history instead of just comparing dependency's history w.r.t. target.
You did not tell make anything about your .o file. Why should make check the timestamp of that file?
Checking the dependency is exactly what make is expected to do.
How should this tool possibly know that a file calles hello_world.o is involved in the process if you don't tell it? There is no magic happening but only the rules of your Makefile are followed.
Is this the limitation of make? How can I circumvent this issue? Because there maybe some external apps messing with the target of my make operation.
You can specify hierarchical dependencies:
all: hello
hello: hello_world.o
<gcc linker command...>
hello_world.o: hello_world.c
gcc hello_world.c -o hello_world
On iOS 5.1.1 the fix for the codesigning has not been applied when jailbreaking. This means that when I launch a ./configure, it fails in its first steps as it seems that the sample programs it run actually cannot be run. There's no patch at this time to fix this.
Making sure that your compiler actually works with no problem, all you can do now is:
trick the ./configure that you are cross compiling by setting different --build and --host triple, such as arm-apple-darwin9 and arm-apple-darwin10;
edit the configure and add BEFORE (eval "$ac_try") 2>&5 this line
(eval "ldid -S ./conftest")
In my experience I only had to do so for the first two checks, the rest is just overkill (as the configure will detect libraries, headers and functions by checking on the compiler result, not on a run test). You should anyway notice that this may not be the only way ./configure tries to run things, and that you will have to run ldid on the final executables before using them (or they will be killed anyway) and that you won't be able to run make check.
Credit: http://botsikas.blogspot.it/2012/06/ipad-ios-51-configure-error-connot-run.html
wait for me to create a C compiler that run ldid automatically (of it's possible). I'm already messing with gcc to make it run on the latest devices such as iPhone 5 (which doesn't accept anymore the arm arch, but only armv7 and newer).
Find a way to tell configure to ldid things (on this you're on your own)
Upon finishing a COBOL program, how do I compile it into an executable file that may be run on other PCs? I'm using OpenCOBOL via cygwin.
Check out this getting started page from the user manual for OpenCOBOL:
But in case the link is broken, just do this:
$ cobc -x hello.cob
$ ./hello
cobc is the compiler. hello.cob is the source file. The output is simply the file hello which can be run by calling ./hello. The -x option is necessary to build an executable.
However, with all compiled programs, it is compiled for the machine is was built on. It will work on machine with similar architectures, but you don't true cross-platform ability unless you're using an interpreted language like Python or Java.
If you compile with Cygwin, the target computers also need Cygwin, or in particular the cygwin dynamic libraries along with the OpenCOBOL runtimes.
Many times, you can also compile under MinGW, which lessens the dependencies, but also lessens the available POSIX features.
Easiest path, install OpenCOBOL and Cygwin on the target machines, and you'll be good to go, otherwise you'll need to produce release packages with all the dependencies and instructions for PATH settings.
I'm trying to build a cross-compiler (x86->64) on my windows system, with the aim of targetting windows 64, however my software currently relies on open source libraries which also have open source dependencies for which there are no prebuilt binaries available with which I can compile. This means that if I want the 64 bit versions I need to compile them.
I've installed MSYS and mingw, I'm also in the process of adding mingw-w64 to the mix so that I can finally compile the libraries in 64 bit form for use with my software. I'm following the steps as closely as I can using these instructions and in the order listed on that page, I'm currently at the step titled "Building the GCC core cross-compiler(s)", but when I try to compile with the line:
$ ../gcc-4.6.1/configure --target=x86_64-w64-mingw32 --enable-targets=all && make -j 6 all-gcc && make -j 6 install-gcc
I get the output pasted here. I should note that I of course snipped the previously executed commands and that last command was the last one listed before all the errors were displayed. Also, I have no idea if it's the cause of all the errors due to the '-j 6' argument, but everything prior to it at least looked successful.
What's the problem and how can I fix it?
Oh, in anticipation of one potential suggestion; no I can't just switch to cygwin.
Edit: Okay after executing them individually, here's the output of the configure command, the output produced by make all-gcc (no -j argument), and config.log. Note, I didn't run a make clean beforehand which may explain the different ending, I didn't do it in the interest of time to write this update, but I suppose I'll just make a different compile folder and re-execute it cleanly to hopefully see the same error as before while I wait for a response.
Edit 2: The make all-gcc failed again as expected, this time the output should help a little more I hope.
Thanks very much for your help.
Your config.log shows that the build process will use the binaries in x86_64-w64-mingw32/bin for stuff like ar, as etc... These are for internal compiler use only, and they should all be available in your /mingw/bin directory. I would strongly suggest asking on the mingw-w64-public mailing list for help.
One one machine, everything compiles fine. On another machine, it complains about the -ly option when I use gcc to create the output file. If I remove the -ly option, then it makes the program, but then it runs differently. What's happening. How can I get the program to run correctly on this linux machine?
You should check to see if you have the same flex/bison versions...
YACC program options (and generated output) vary from OS to OS. Bison might be more consistent.