Hi I have been working on the proposal of this project, but still have some questions, could anyone please give me some help? For the bits part, what does the 'compressed' mean, what is the main difference between the array, vector and the normal ones? Is there any other existing integral special functions that mentioned in the second part? Where can I find them? Thanks a lot Shiyang Yang
For the bits part, what does the 'compressed' mean, what is the main difference between the array, vector and the normal ones?
Compressed is probably a poor choice of wording. It should have been "compact". Essentially, this means that the objects in the vector or array are encoded by the minimum bit-width (e.g., vector<bool>).
Is there any other existing integral special functions that mentioned in the second part? Where can I find them?
I'm not aware of any. Maybe Jeremiah can provide some additional references? Andrew Sutton andrew.n.sutton@gmail.com
On Thu, 8 Apr 2010, Andrew Sutton wrote:
Is there any other existing integral special functions that mentioned in the second part? Where can I find them?
I'm not aware of any. Maybe Jeremiah can provide some additional references?
Most of the stuff is from the URL that is on the Wiki page; a lot of the stuff is fairly basic but part of the goal is to abstract special processor instructions that do the operations with the best performance (rotate instructions, conditional moves based on the carry bit, etc.). That covers items 1 and 7 from the second bulleted list under Bits & Ints. Sign extension and saturating arithmetic may also have special instructions, as might bit-reversed addition and similar operations, but portable versions would also be necessary for platforms that don't have special instructions. Some more references are: Chapter 8 of http://support.amd.com/us/Processor_TechDocs/40546-PUB-Optguide_3-11_5-21-09... (for integer division by a repeated value) http://graphics.stanford.edu/~seander/bithacks.html (we probably should have everything in here) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.86.1801 (with the addition routines for interleaved integers using arbitrary masks) A couple of new things to add to the list: generic versions of count-leading-zeros, count-trailing-zeros, population count, etc. that have gcc intrinsics (only for certain concrete types, and with different names for different types) and often special instructions. -- Jeremiah Willcock
participants (3)
-
Andrew Sutton -
Jeremiah Willcock -
shiyang yang