[Lame-dev] [ lame-Bugs-3025801 ] gain_analysis.h needs int constants

[SourceForge.net <no...@so...>]

Bugs item #3025801, was opened at 2010-07-06 11:39
Message generated for change (Tracker Item Submitted) made by nobody
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=100290&aid=3025801&group_id=290

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
Summary: gain_analysis.h needs int constants

Initial Comment:
The MinGW version of GCC 4.5.0, compiling the command-line LAME
encoder, complains about variably defined A and B arrays in
gain_analysis.h.  The problem comes from using these floating-point
constants to define array sizes:

#define STEPS_per_dB      100. /* Table entries per dB */
#define MAX_dB            120. /* Table entries for 0...MAX_dB (normal max. values are 70...80 dB) */

I added these two lines, immediately after the two above.
Note the lack of decimal points.

#define INT_STEPS_per_dB  100 /* Table entries per dB */
#define INT_MAX_dB        120 /* Table entries for 0...MAX_dB (normal max. values are 70...80 dB) */

Then I changed the A and B declarations to use the new integer
constants, and the compiler was happy.

        uint32_t A[(size_t) (INT_STEPS_per_dB * INT_MAX_dB)];
        uint32_t B[(size_t) (INT_STEPS_per_dB * INT_MAX_dB)];

The (size_t) cast is probably now redundant, but I left it in to
make as few changes as possible.


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=100290&aid=3025801&group_id=290