• Advertisement
Sign in to follow this  

Bit Flags vs. Boolean

This topic is 1925 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I was trying to explain to a semi-programmer friend of mine today the purpose of Bit Flags vs. Booleans. And I couldn't think of any besides Memory Usage which seems rather weak seeing as it would save you a few 100 bytes of memory at most. So what's a could argument for Bit Flags I can tell him about, and to be honest I'm wondering myself right now too.

Share this post


Link to post
Share on other sites
Advertisement
It tends to be a bit more memory than that in most use cases. You often don't just have one bit flag, you have one bit flag per user/entity/product/etc. It adds up.

It also speeds certain operations since an equality check is 1 op rather than N. And/Or/Xor/Negate are also 1 op instead of N.

Reading and writing from disk/database/network is faster since there's less space and/or compression going on.

I'm sure there's a few others I'm missing.

Share this post


Link to post
Share on other sites
One example of a performance benefit would be doing something like render queue sorting based on a bitfield. It would be much better to sort based on a single field rather than a bunch of single flags.

Share this post


Link to post
Share on other sites
Bitfields were available in C because C was used to develop the Unix operating system, and operating systems deal with hardware. Hardware control registers often have bit-addressable functions.

C also did not have a boolean type. The PDP architeture it was first developed on has a (bit-addressable) flag register in which the Z flag gets set every time an operation yields a zero result (and cleared otherwise), so it was easier to have control-flow operations act on zero or non-zero integral values. BTW, that's also why null-terminated character strings were used instead of the more-common (at the time) Pascal-style counted strings.

Share this post


Link to post
Share on other sites
One particular application you might want to look at is [url=http://en.wikipedia.org/wiki/Bitboard]Bitboards[/url].

Share this post


Link to post
Share on other sites
[quote name='SiCrane' timestamp='1349925185' post='4988945']
One particular application you might want to look at is Bitboards.
[/quote]
This is probably the best example and is also what I was hinting at by using the bitboard concept to sort render ops via state, etc.

Share this post


Link to post
Share on other sites
One thing to keep in mind is that accessing bit fields will require the compiler to produce additional bit masking instructions, which will *increase* the memory footprint of your instruction code. If you try to use bit fields as a general rule instead of a situational optimisation, you'll end up with slower code.

Also, the alignment issues are subtle. Most types want to align on 4 byte boundaries, so AFAIK if you have a structure containing, say, an integer, and any less than 5 flags, then converting these flags to single bits won't save any space at all.

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement