Source: insighttoolkit4 Version: 4.13.3withdata-dfsg1-3 Severity: normal
I'm sure there is a good reason for the remove_gcc_version_test.patch which removes a load of variable setting by GCC presence and version in Modules/ThirdParty/VNL/src/vxl/vcl/vcl_compiler.h However, doing this causes VCL_GCC not to be defined which causes this snippet further down the file to use the definition with 'export' on the end: // if the compiler doesn't understand "export", we just leave it out. // gcc understands it, but ignore it noisily. // icc understands it, but ignore it noisily. #if !VCL_HAS_EXPORT || defined(VCL_GCC) || defined(VCL_ICC) # define VCL_TEMPLATE_EXPORT /* ignore */ #else # define VCL_TEMPLATE_EXPORT export #endif Every time that define is used we get pointless warning noise of the form; insighttoolkit4/Modules/ThirdParty/VNL/src/vxl/vcl/vcl_compiler.h:114:30: warning: keyword 'export' not implemented, and will be ignored That wouldn't matter too much except that it occurs 31,024 times (!) during the build, which really does distract from things which might actually be real issues and bloats the log. Removing that patch so that VCL_GCC gets set makes this problem go away. So, is that patch still useful, and why? Can we please at least leave this bit: #if defined(__GNUC__) && !defined(__ICC) // icc 8.0 defines __GNUC__ # define VCL_GCC #endif to kill the epic level of 'export' whinges? -- Wookey
