Will you please switch to iterators here?

LGTM, thanks!

LGTM, thanks!

Looks good to me! FP was an old register name from mspgcc time. MSP430 EABI does not use frame pointer at all (seems to be a bug in the ABI, as I do not understand how we could support VLAs or alloca's in this case).

In general LGTM. But will you please correct the test names, so I can commit on your behalf?

@joanlluch Yes, having the version with amount does make sense as target may know how to efficiently lower some of the shifts with crazy tricks. And in the view of this patch it's easier to preserve the sequences rather than match and undo later. I guess for MSP430 we'd need to allow 9 as possible "special" shift value, as select is lowered to 3 instructions (test + branch + mov of some kind) and removing branch overall has some additional value like allowing more code folding / reducing register pressure.

LGTM, thanks!

If they are not necessary, then it would make sense to undo them. At least partially.

@joanlluch There are two things here. We understand that there might be some out-of-tree targets that would benefit from changes in the mainline. However, it is still important to add tests that utilize mainline targets as this ensures that the proposed changes won't e.g. break during some refactoring, etc. So, using MSP430 (probably the smallest) / AVR is perfectly fine for tests, etc.!

The MSP430 changes / overall direction looks extremely useful to me. We just need to ensure everything is done properly. This is exactly what @spatel proposed.

LGTM, thanks!

The changes are pretty straightforward :)

Note that on AArch64, there is no equivalent to x86's use of "al" for varargs calls.

Correct. Everywhere else we're using the same ABI regardless of this flag.

@efriedma The LangRef states that "This attributes disables implicit floating-point instructions". I believe that the definition of "implicit" is very vague here. And clang cmdline documentation does not explain "-mno-implicit-float" at all. What you'd suggest? :)

How many such switches are seen in e.g. LLVM test suite? Could you please share the size reduction statistics?

Well, what's the overall motivation of all these changes? There are no multiple threads, etc. on msp430. I believe instead of lowering to libcalls we could simply "drop" atomic.

@wuzish Do you really wanted on comment on commit that is more than 10 years old? Have you looked over the current ToT?

Paul, sorry, I completely missed this review request for some reason... Please rebase and commit, thanks!

Ok now

LGTM now

LGTM

Please make sure the test passes.

The patch was reverted – the test requires fully configured msp430 configuration and therefore was breaking the bots.

Patch fails to apply. Please rebase.

This breaks CodeGen/MSP430/mult-alt-generic-msp430.ll :)