Jochen Kilian
2010-Aug-06 09:56 UTC
[CELT-dev] Issues to build fixed point version of celt 0.8.1 under windows
Hi All I am trying to build the latest version 0.8.1 using Cygwin under Windows. Configure: Everything builds well using the default configuration (configure with no additional parameters). trying to use: configure --enable-fixed results in a warning: unrecognized option. As I found --enable-fixed-point in the config.in as parameter I run the configure with --enable-fixed-point Now the configure runs well. The final screen of the configure says: Automatic configuration OK Fixed point support <yes> Make: But if I run now make, I do end up with the following error during linkage: in libcelt/mathops.h:176: undefined reference to `_ec_ilog? There are also many warnings with: visibility attribute not supported for this configuration. Any ideas? Was somebody able to build the fixed point version? In the same context if I want to evaluate the quality: What is the quality difference between the fixed and the floating point version? Thank you very much Jochen
Jochen Kilian
2010-Aug-06 10:03 UTC
[CELT-dev] Issues to build fixed point version of celt 0.8.1 under windows
Hi All I am trying to build the latest version 0.8.1 using Cygwin under Windows. Configure: Everything builds well using the default configuration (configure with no additional parameters). trying to use: configure --enable-fixed results in a warning: unrecognized option. As I found --enable-fixed-point in the config.in as parameter I run the configure with --enable-fixed-point Now the configure runs well. The final screen of the configure says: Automatic configuration OK Fixed point support <yes> Make: But if I run now make, I do end up with the following error during linkage: in libcelt/mathops.h:176: undefined reference to `_ec_ilog? There are also many warnings with: visibility attribute not supported for this configuration. Any ideas? Was somebody able to build the fixed point version? In the same context if I want to evaluate the quality: What is the quality difference between the fixed and the floating point version? Thank you very much Jochen