This is the mail archive of the newlib@sourceware.org mailing list for the newlib project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
This change results in the function working two times faster on ARM Cortex-M3 microcontroller - one million calls of "old" function using random numbers take ~70s, while the same calculations with "new" code take ~33s (clock of chip is 16MHz). Obviously the improvement depends on the input value - sometimes the patched code can be slower than the old one - for input values "close" to year 1970. On the other hand, the "further" from 1970, the bigger is the performance gain.
Both versions give the same results - tested with a few million random numbers - both versions were called for the same input value and result was compared with memcmp().
Regards, FCh
Attachment:
ChangeLog.txt
Description: Text document
Attachment:
0001-_mktm_r-optimize-speed.patch
Description: Text document
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |