Hello, MIPSpro tests haven't been updated since Aug 18, yet I see some recent fixes being done in the CVS for this compiler. Is there anywhere else where tests for this platform can be consulted? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
--- Joaqu�n M� L�pez Mu�oz <joaquin@tid.es> wrote:
MIPSpro tests haven't been updated since Aug 18, yet I see some recent fixes being done in the CVS for this compiler.
We fixed the MIPSpro 7.3 problems only yesterday.
Is there anywhere else where tests for this platform can be consulted?
I just started a new run. It will take 5-6 hours before the results show up at the Metacomm site. Ralf __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Ralf W. Grosse-Kunstleve wrote:
--- Joaquín Mª López Muñoz <joaquin@tid.es> wrote:
MIPSpro tests haven't been updated since Aug 18, yet I see some recent fixes being done in the CVS for this compiler.
We fixed the MIPSpro 7.3 problems only yesterday.
I ran the regression tests for mipspro 7.4 in the past. But since this compiler does not evolve at all anymore and the hardware-line is end-of-life we are trying to switch to gcc on this platform: porting boost to this platform is too much of a hassle without proper support from the vendor.
--- Toon Knapen <toon.knapen@fft.be> wrote:
Ralf W. Grosse-Kunstleve wrote:
--- Joaqu�n M� L�pez Mu�oz <joaquin@tid.es> wrote:
MIPSpro tests haven't been updated since Aug 18, yet I see some recent fixes being done in the CVS for this compiler.
We fixed the MIPSpro 7.3 problems only yesterday.
I ran the regression tests for mipspro 7.4 in the past. But since this compiler does not evolve at all anymore and the hardware-line is end-of-life we are trying to switch to gcc on this platform: porting boost to this platform is too much of a hassle without proper support from the vendor.
FWIW: Boost.Python works with both MIPSpro 7.3 and 7.4. The only test that fails is the embedding test. MIPSpro 7.4 == EDG 245, a very good front end. It shouldn't really be a problem to make most things work with this compiler, although it will of course require some time to work out the fixes. MIPSpro 7.3 == EDG 238 is a different story, though, quite a nightmare (but not nearly as bad as VC6 which is almost triply obsolete now (7,7.1,8.0)). Ralf __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail
Ralf W. Grosse-Kunstleve wrote:
FWIW: Boost.Python works with both MIPSpro 7.3 and 7.4. The only test that fails is the embedding test. MIPSpro 7.4 == EDG 245, a very good front end. It shouldn't really be a problem to make most things work with this compiler, although it will of course require some time to work out the fixes.
I'm prepared to implement fixes for IBM and HP because I know that the fix is temporary. The compiler evolves and in the future the fix will not be necessary anymore. So implementing a fix is just allowing to use the functionality already before the compiler is capable to of compiling the standard-compliant implementation of boost. But the mipspro compiler does not evolve anymore so you know in advance that will be stuck fixing things the whole time without the compiler catching up. And additionally I suspect that in 2 years the platform will not be used anymore so why bother.
--- Toon Knapen <toon.knapen@fft.be> wrote:
Ralf W. Grosse-Kunstleve wrote:
FWIW: Boost.Python works with both MIPSpro 7.3 and 7.4. The only test that fails is the embedding test. MIPSpro 7.4 == EDG 245, a very good front end. It shouldn't really be a problem to make most things work with this compiler, although it will of course require some time to work out the fixes.
I'm prepared to implement fixes for IBM and HP because I know that the fix is temporary. The compiler evolves and in the future the fix will not be necessary anymore. So implementing a fix is just allowing to use the functionality already before the compiler is capable to of compiling the standard-compliant implementation of boost. But the mipspro compiler does not evolve anymore so you know in advance that will be stuck fixing things the whole time without the compiler catching up. And additionally I suspect that in 2 years the platform will not be used anymore so why bother.
It is still important to us, unfortunately. I can see that it is not worth bothering for you, but your first email may have lead some people to the impression that MIPSpro doesn't work at all for Boost, which is clearly not the case. I just wanted to make sure people know the MIPSpro can be made to work, as Boost.Python demonstrates. Ralf __________________________________ Do you Yahoo!? Yahoo! Mail - Helps protect you from nasty viruses. http://promotions.yahoo.com/new_mail
Ralf W. Grosse-Kunstleve wrote:
It is still important to us, unfortunately. I can see that it is not worth bothering for you, but your first email may have lead some people to the impression that MIPSpro doesn't work at all for Boost, which is clearly not the case. I just wanted to make sure people know the MIPSpro can be made to work, as Boost.Python demonstrates. Ralf
Sorry if I gave the impression that boost does not work on SGI, I never meant to imply that. I also mentioned that we were planning to switch to gcc on SGI. However gcc on SGI does seem to have its own set of problems. I therefore think that we will nevertheless continue to run the regressions tests on SGI using the mipspro compiler.
participants (6)
-
Joaquín Mª López Muñoz -
Ralf W. Grosse-Kunstleve -
Toon Knapen