small patch for translator/c/src/ll_osdefs.h for win32
I attempted to translate a pypy-c in windows today. I had to mask out including sys/mman.h. Afterwards, I was able to succesfully translate. Index: ll_osdefs.h =================================================================== --- ll_osdefs.h (revision 30390) +++ ll_osdefs.h (working copy) @@ -42,7 +42,9 @@ #include <sys/types.h> +#if !defined(MS_WINDOWS) #include <sys/mman.h> +#endif #ifdef HAVE_STROPTS_H #include <stropts.h> -- Scott Dial scott@scottdial.com scodial@indiana.edu
On 7/24/06, Scott Dial <scott+pypy-dev@scottdial.com> wrote:
That was my fault. It has been fixed. Thanks -- Lawrence http://www.oluyede.org/blog
Hi all, On Tue, Jul 25, 2006 at 06:45:17AM +0200, Lawrence Oluyede wrote:
Now that we have automated test runs and translations on Linux machines, we could try to think how we could go about doing the same on Windows machines. Very few of us actually have Windows machines around, so it would be important, otherwise things will break more and more (it's already a good surprize that a PyPy translation still works after just a quick fix :-) """What is not tested is broken.""" A bientot, Armin
Armin Rigo wrote:
I actually was pretty suprised too.. Presuming this is an important platform to test (*grin*), I can provide a daily testing run for Windows. I don't know what the code behind your pypytest is, but I presume I could run something similar? Let me know. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scottt, On Tue, Jul 25, 2006 at 07:44:45AM -0400, Scott Dial wrote:
The scripts I use in the following directory: http://codespeak.net/svn/user/arigo/hack/misc/htmlconftest/ While some details might be Unix-specific, it shouldn't be too hard to adapt them to Windows. (They produce symlinks, for example.) Put the scripts in a directory, together with a subdir 'html' (initially empty) which is visible on the web, and a checkout of PyPy in 'pypy-dist'. The entry point script is autotest.py. A bientot, Armin
Armin Rigo wrote:
While some details might be Unix-specific, it shouldn't be too hard to adapt them to Windows. (They produce symlinks, for example.)
I was able to adapt the script just fine. And I ran it once without problems (with the script, not pypy ;). But, I realized I made an error and forget to setup the build enviroment for compiling. After doing so, I find that I get a stalled in module\_stackless\test Nevertheless, in the future the results will be at http://scottdial.com/pypytest/ I will look into patching the misbehaving tests. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott, On Fri, Jul 28, 2006 at 01:41:11PM -0400, Scott Dial wrote:
Nevertheless, in the future the results will be at http://scottdial.com/pypytest/
Thanks a lot! Can we link to this page from the "Status" section of the documentation index? A bientot, Armin
Armin Rigo wrote:
You sure can. I'll do my best to keep it available as much as possible. I'll be moving in 3 weeks and there'll be some days of downtime while I get my internet hooked up there, but otherwise I intend to keep it updated and available. I'm not sure what time is the best for it to run daily. I've currently set it run daily at midnight GMT. Although I may have to change that depending on my sleeping behavior.. when it runs probably isn't that big of a concern as long it gets run, yah? -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott, On Sat, Jul 29, 2006 at 11:27:49AM -0400, Scott Dial wrote:
You sure can.
Thanks!
Indeed. On snake it's in the early morning, just to minimize the risk that someone else is doing something. A bientot, Armin
Scott Dial wrote:
I'm not sure what time is the best for it to run daily. I've currently set it run daily at midnight GMT.
For the record, I have a need for this to happen much later. I have moved it to 9am GMT which I realize may be afternoon for some people in Europe, but I can't help the Earth's rotation determining my schedule. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott! Scott Dial wrote:
Just for the record, I think we all appreciate very much that you are running the tests and it does not matter too much _when_ you run them :-). Thank you! Cheers, Carl Friedrich
Scott Dial <scott+pypy-dev@scottdial.com> writes:
How could 9am _Greenwich_ (a place in the UK) Mean Time be in the afternoon anywhere in Europe? :)
but I can't help the Earth's rotation determining my schedule.
Indeed :) As Carl says, thanks a lot for doing this! Cheers, mwh -- [1]For those of you who aren't aware "tossing" is a euphamism for, well, vigourously rubbing your love pole. You understand? Flogging the dolphin. Stretching the chicken's neck. Waving your magic wand. Basically, wanking. -- Just another Morfans SDA update
Michael Hudson wrote:
How could 9am _Greenwich_ (a place in the UK) Mean Time be in the afternoon anywhere in Europe? :)
*blush* I should avoid writing emails at 2am (localtime). :-) -- Scott Dial scott@scottdial.com scodial@indiana.edu
I think it could also be better to refactor a bit that header. I placed the header I needed for my extension modules but I realize that's not the right place for them. We can move out "my" includes in a separate modules.h and include it in g_prerequisite.h or g_include.h or whatever. -- Lawrence http://www.oluyede.org/blog
Hi Lawrence, On Thu, Jul 27, 2006 at 11:17:12AM +0200, Lawrence Oluyede wrote:
The rctypes-based modules' includes shouldn't go in the static .h files. They are generated automatically by genc if you use the correct magic: look for _includes_ in module/readline/c_readline.py. A bientot, Armin
I definitely missed that! Thanks. -- Lawrence http://www.oluyede.org/blog
On 7/27/06, Lawrence Oluyede <l.oluyede@gmail.com> wrote:
That doesn't seem to work. I commented out the includes I added to make fcntl module compile (like sys/ioctl.h for example) and added them to the _includes_ and this seem to work correctly only with tests and the generation of ctypesplatcheck files. The headers are not injected in the generated fcntl.c nor in any of the file it includes. -- Lawrence http://www.oluyede.org/blog
On 7/27/06, Lawrence Oluyede <l.oluyede@gmail.com> wrote:
That doesn't seem to work. I commented out the includes I added to make fcntl module compile (like sys/ioctl.h for example) and added them to the _includes_ and this seem to work correctly only with tests and the generation of ctypesplatcheck files. The headers are not injected in the generated fcntl.c nor in any of the file it includes. -- Lawrence http://www.oluyede.org/blog
Hi Lawrence, On Thu, Jul 27, 2006 at 01:32:25PM +0200, Lawrence Oluyede wrote:
Strange: it works with the readline module. If you run 'compilemodule.py readline' you can check that /tmp/usession-*/readline/common_header.h contains '#include <readline/readline.h>'. A bientot, Armin.
On 7/24/06, Scott Dial <scott+pypy-dev@scottdial.com> wrote:
That was my fault. It has been fixed. Thanks -- Lawrence http://www.oluyede.org/blog
Hi all, On Tue, Jul 25, 2006 at 06:45:17AM +0200, Lawrence Oluyede wrote:
Now that we have automated test runs and translations on Linux machines, we could try to think how we could go about doing the same on Windows machines. Very few of us actually have Windows machines around, so it would be important, otherwise things will break more and more (it's already a good surprize that a PyPy translation still works after just a quick fix :-) """What is not tested is broken.""" A bientot, Armin
Armin Rigo wrote:
I actually was pretty suprised too.. Presuming this is an important platform to test (*grin*), I can provide a daily testing run for Windows. I don't know what the code behind your pypytest is, but I presume I could run something similar? Let me know. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scottt, On Tue, Jul 25, 2006 at 07:44:45AM -0400, Scott Dial wrote:
The scripts I use in the following directory: http://codespeak.net/svn/user/arigo/hack/misc/htmlconftest/ While some details might be Unix-specific, it shouldn't be too hard to adapt them to Windows. (They produce symlinks, for example.) Put the scripts in a directory, together with a subdir 'html' (initially empty) which is visible on the web, and a checkout of PyPy in 'pypy-dist'. The entry point script is autotest.py. A bientot, Armin
Armin Rigo wrote:
While some details might be Unix-specific, it shouldn't be too hard to adapt them to Windows. (They produce symlinks, for example.)
I was able to adapt the script just fine. And I ran it once without problems (with the script, not pypy ;). But, I realized I made an error and forget to setup the build enviroment for compiling. After doing so, I find that I get a stalled in module\_stackless\test Nevertheless, in the future the results will be at http://scottdial.com/pypytest/ I will look into patching the misbehaving tests. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott, On Fri, Jul 28, 2006 at 01:41:11PM -0400, Scott Dial wrote:
Nevertheless, in the future the results will be at http://scottdial.com/pypytest/
Thanks a lot! Can we link to this page from the "Status" section of the documentation index? A bientot, Armin
Armin Rigo wrote:
You sure can. I'll do my best to keep it available as much as possible. I'll be moving in 3 weeks and there'll be some days of downtime while I get my internet hooked up there, but otherwise I intend to keep it updated and available. I'm not sure what time is the best for it to run daily. I've currently set it run daily at midnight GMT. Although I may have to change that depending on my sleeping behavior.. when it runs probably isn't that big of a concern as long it gets run, yah? -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott, On Sat, Jul 29, 2006 at 11:27:49AM -0400, Scott Dial wrote:
You sure can.
Thanks!
Indeed. On snake it's in the early morning, just to minimize the risk that someone else is doing something. A bientot, Armin
Scott Dial wrote:
I'm not sure what time is the best for it to run daily. I've currently set it run daily at midnight GMT.
For the record, I have a need for this to happen much later. I have moved it to 9am GMT which I realize may be afternoon for some people in Europe, but I can't help the Earth's rotation determining my schedule. -- Scott Dial scott@scottdial.com scodial@indiana.edu
Hi Scott! Scott Dial wrote:
Just for the record, I think we all appreciate very much that you are running the tests and it does not matter too much _when_ you run them :-). Thank you! Cheers, Carl Friedrich
Scott Dial <scott+pypy-dev@scottdial.com> writes:
How could 9am _Greenwich_ (a place in the UK) Mean Time be in the afternoon anywhere in Europe? :)
but I can't help the Earth's rotation determining my schedule.
Indeed :) As Carl says, thanks a lot for doing this! Cheers, mwh -- [1]For those of you who aren't aware "tossing" is a euphamism for, well, vigourously rubbing your love pole. You understand? Flogging the dolphin. Stretching the chicken's neck. Waving your magic wand. Basically, wanking. -- Just another Morfans SDA update
Michael Hudson wrote:
How could 9am _Greenwich_ (a place in the UK) Mean Time be in the afternoon anywhere in Europe? :)
*blush* I should avoid writing emails at 2am (localtime). :-) -- Scott Dial scott@scottdial.com scodial@indiana.edu
I think it could also be better to refactor a bit that header. I placed the header I needed for my extension modules but I realize that's not the right place for them. We can move out "my" includes in a separate modules.h and include it in g_prerequisite.h or g_include.h or whatever. -- Lawrence http://www.oluyede.org/blog
Hi Lawrence, On Thu, Jul 27, 2006 at 11:17:12AM +0200, Lawrence Oluyede wrote:
The rctypes-based modules' includes shouldn't go in the static .h files. They are generated automatically by genc if you use the correct magic: look for _includes_ in module/readline/c_readline.py. A bientot, Armin
I definitely missed that! Thanks. -- Lawrence http://www.oluyede.org/blog
On 7/27/06, Lawrence Oluyede <l.oluyede@gmail.com> wrote:
That doesn't seem to work. I commented out the includes I added to make fcntl module compile (like sys/ioctl.h for example) and added them to the _includes_ and this seem to work correctly only with tests and the generation of ctypesplatcheck files. The headers are not injected in the generated fcntl.c nor in any of the file it includes. -- Lawrence http://www.oluyede.org/blog
On 7/27/06, Lawrence Oluyede <l.oluyede@gmail.com> wrote:
That doesn't seem to work. I commented out the includes I added to make fcntl module compile (like sys/ioctl.h for example) and added them to the _includes_ and this seem to work correctly only with tests and the generation of ctypesplatcheck files. The headers are not injected in the generated fcntl.c nor in any of the file it includes. -- Lawrence http://www.oluyede.org/blog
Hi Lawrence, On Thu, Jul 27, 2006 at 01:32:25PM +0200, Lawrence Oluyede wrote:
Strange: it works with the readline module. If you run 'compilemodule.py readline' you can check that /tmp/usession-*/readline/common_header.h contains '#include <readline/readline.h>'. A bientot, Armin.
participants (5)
-
Armin Rigo -
Carl Friedrich Bolz -
Lawrence Oluyede -
Michael Hudson -
Scott Dial