build: Rename --enable-experimental-asm to --enable-asm and enable by default

[laanwj]

Now that 0.15 is branched off, enable assembler SHA256 optimizations by default, but still allow disabling them, for example if something goes wrong with auto-detection on a platform.

Also add mention of the use of asm in the configure summary.

[promag]

Should the flag mention SHA256 like --enable-asm-sha256?

Either way LGTM.

[laanwj]

Should the flag mention SHA256 like --enable-asm-sha256?

No, I think we should have a general asm flag, build-side, as this does. This is similar to OpenSSL's flag to use assembly or not. Something else creates too many combinations to test, and I don't see why it would be useful.
(besides testing, but run-time selection is much more useful for that)

[promag]

Ok, just had to ask.

utACK 538cc0c.

[kallewoof]

If this is meant for 0.15, the release notes need to be updated. They currently read

SHA256 hashing has been optimized for architectures supporting SSE 4 (See PR 10182). SHA256 is around 50% faster on supported hardware, which results in around 5% faster IBD and block validation. In version 0.15, SHA256 hardware optimization is disabled in release builds by default, but can be enabled by using --enable-experimental-asm when building.

[laanwj]

No, this is not meant for 0.15, certainly not for 0.15.0

Now that 0.15 is branched off, enable assembler SHA256 optimizations by default

[kallewoof]

Sorry about that. I should've read the PR message more closely.

[sipa]

@laanwj There are plenty more versions of SHA256 to add (for starters, a RORX version, a SHA-NI version, but later perhaps a 4-way parallel SSE4 or 8-way parallel AVX2 version). Do we all want to include them under "--enable-asm", or will they first go into a reinstated "--enable-experimental-asm" before being promoted?

[laanwj]

Do we all want to include them under "--enable-asm" or will they first go into a reinstated "--enable-experimental-asm" before being promoted?

Both are valid options, we should probably decide on a case by case basis - depending on how much we trust the code, how long it can be in master until the release.

[luke-jr]

utACK

[fanquake]

utACK 538cc0c

[droark]

Posthumous review comment: Is USE_ASM meant to be a catch-all for any other assembly optimizations added in the future, or is it solely for the SHA256 optimizations? I ask because the language in the changes is a little ambiguous (IMO). I'd like to fix that. I just don't want to step on toes first.

Thanks.