91 Commits

Author SHA1 Message Date
Ayke van Laethem
9592920890
[AVR] Optimize 32-bit shifts: optimize REG_SEQUENCE
This pseudo-instruction stores two small (8-bit) registers into one wide
(16-bit) register. But apparently the order matters a lot to the
register allocator.
This patch changes the order of inserting the registers to optimize for
the best register allocation in the tests of shift32.ll. It might be
detrimental in other cases, but keeping the registers in the same
physical register seems like it would be a common case.

Differential Revision: https://reviews.llvm.org/D140573
2023-01-08 20:05:31 +01:00
Ayke van Laethem
fad5e0cf50
[AVR] Optimize 32-bit shifts: reverse shift + move
This optimization turns shifts of almost a multiple of 8 into a shift
into the opposite direction. Unfortunately it doesn't compose well with
the other optimizations (I've tried) so it's separate from them.

Differential Revision: https://reviews.llvm.org/D140572
2023-01-08 20:05:31 +01:00
Ayke van Laethem
81f5f22f27
[AVR] Optimize 32-bit shifts: shift by 4 bits
This uses a complicated shift sequence that avr-gcc also uses, but
extended to work over any number of bytes and in both directions
(logical shift left and logical shift right). Unfortunately it can't be
used for an arithmetic shift right: I've tried to come up with a
sequence but couldn't.

Differential Revision: https://reviews.llvm.org/D140571
2023-01-08 20:05:31 +01:00
Ayke van Laethem
8f8afabd32
[AVR] Optimize 32-bit shift: move bytes around
This patch optimizes 32-bit constant shifts by renaming registers. This
is very effective as the compiler would otherwise need to do a lot of
single bit shift instructions. Instead, the registers are renamed at the
SSA level which means the register allocator will insert the necessary
mov instructions.

Unfortunately, the register allocator will insert some unnecessary movs
with the current code. This will be fixed in a later patch.

Differential Revision: https://reviews.llvm.org/D140570
2023-01-08 20:05:31 +01:00
Ayke van Laethem
840d10a1d2
[AVR] Custom lower 32-bit shift instructions
32-bit shift instructions were previously expanded using the default
SelectionDAG expander, which meant it used 16-bit constant shifts and
ORed them together. This works, but is far from optimal.

I've optimized 32-bit shifts on AVR using a custom inserter. This is
done using three new pseudo-instructions that take the upper and lower
bits of the value in two separate 16-bit registers and outputs two
16-bit registers.

This is the first commit in a series. When completed, shift instructions
will take around 31% less instructions on average for constant 32-bit
shifts, and is in all cases equal or better than the old behavior. It
also tends to match or outperform avr-gcc: the only cases where avr-gcc
does better is when it uses a loop to shift, or when the LLVM register
allocator inserts some unnecessary movs. But it even outperforms avr-gcc
in some cases where avr-gcc does not use a loop.

As a side effect, non-constant 32-bit shifts also become more efficient.

For some real-world differences: the build of compiler-rt I use in
TinyGo becomes 2.7% smaller and the build of picolibc I use becomes 0.9%
smaller. I think picolibc is a better representation of real-world code,
but even a ~1% reduction in code size is really significant.

The current patch just lays the groundwork. The result is actually a
regression in code size. Later patches will use this as a basis to
optimize these shift instructions.

Differential Revision: https://reviews.llvm.org/D140569
2023-01-08 20:05:31 +01:00
serge-sans-paille
38818b60c5
Move from llvm::makeArrayRef to ArrayRef deduction guides - llvm/ part
Use deduction guides instead of helper functions.

The only non-automatic changes have been:

1. ArrayRef(some_uint8_pointer, 0) needs to be changed into ArrayRef(some_uint8_pointer, (size_t)0) to avoid an ambiguous call with ArrayRef((uint8_t*), (uint8_t*))
2. CVSymbol sym(makeArrayRef(symStorage)); needed to be rewritten as CVSymbol sym{ArrayRef(symStorage)}; otherwise the compiler is confused and thinks we have a (bad) function prototype. There was a few similar situation across the codebase.
3. ADL doesn't seem to work the same for deduction-guides and functions, so at some point the llvm namespace must be explicitly stated.
4. The "reference mode" of makeArrayRef(ArrayRef<T> &) that acts as no-op is not supported (a constructor cannot achieve that).

Per reviewers' comment, some useless makeArrayRef have been removed in the process.

This is a follow-up to https://reviews.llvm.org/D140896 that introduced
the deduction guides.

Differential Revision: https://reviews.llvm.org/D140955
2023-01-05 14:11:08 +01:00
Ayke van Laethem
5527b21516
[AVR] Do not use R0/R1 on avrtiny
This patch makes sure the compiler uses R16/R17 on avrtiny (attiny10
etc) instead of R0/R1.

Some notes:

  * For the NEGW and ROLB instructions, it adds an explicit zero
    register. This is necessary because the zero register is different
    on avrtiny (and InstrInfo Uses lines need a fixed register).
  * Not entirely sure about putting all tests in features/avr-tiny.ll,
    but it doesn't seem like the "target-cpu"="attiny10" attribute
    works.

Updates: https://github.com/llvm/llvm-project/issues/53459

Differential Revision: https://reviews.llvm.org/D138582
2022-11-28 18:05:55 +01:00
Simon Pilgrim
51d5fd4321 Fix MSVC "signed/unsigned mismatch" warning. NFC. 2022-11-24 14:08:40 +00:00
Ben Shi
f452b9dcaf [AVR] Fix wrong ABI of AVRTiny.
A scalar which exceeds 4 bytes should be returned via stack, other
than via registers, on an AVRTiny device.

Reviewed By: aykevl

Differential Revision: https://reviews.llvm.org/D138201
2022-11-23 09:32:47 +08:00
Sergei Barannikov
c6acb4eb0f [SDAG] Add getCALLSEQ_END overload taking uint64_ts
All in-tree targets pass pointer-sized ConstantSDNodes to the
method. This overload reduced amount of boilerplate code a bit.  This
also makes getCALLSEQ_END consistent with getCALLSEQ_START, which
already takes uint64_ts.
2022-09-15 14:02:12 -04:00
Joe Loser
5e96cea1db [llvm] Use std::size instead of llvm::array_lengthof
LLVM contains a helpful function for getting the size of a C-style
array: `llvm::array_lengthof`. This is useful prior to C++17, but not as
helpful for C++17 or later: `std::size` already has support for C-style
arrays.

Change call sites to use `std::size` instead.

Differential Revision: https://reviews.llvm.org/D133429
2022-09-08 09:01:53 -06:00
Ayke van Laethem
a560e57a7e
[AVR] Only push and clear R1 in interrupts when necessary
R1 is a reserved register, but LLVM gives the APIs to know when it is
used or not. So this patch uses these APIs to only save/clear/restore R1
in interrupts when necessary.

The main issue here was getting inline assembly to work. One could argue
that this is the job of Clang, but for consistency I've made sure that
R1 is always usable in inline assembly even if that means clearing it
when it might not be needed.

Information on inline assembly in AVR can be found here:

https://www.nongnu.org/avr-libc/user-manual/inline_asm.html#asm_code

Essentially, this seems to suggest that r1 can be freely used in avr-gcc
inline assembly, even without specifying it as an input operand.

Differential Revision: https://reviews.llvm.org/D117426
2022-08-15 14:29:38 +02:00
Ben Shi
bce2e208e0 [AVR] Optimize int16 airthmetic right shift for shift amount 7/14/15
Reviewed By: aykevl

Differential Revision: https://reviews.llvm.org/D115618
2022-03-26 06:53:27 +00:00
Ben Shi
f319c24570 [AVR] Reject/Reserve R0~R15 on AVRTiny.
Reviewed By: aykevl, dylanmckay

Differential Revision: https://reviews.llvm.org/D121672
2022-03-24 02:33:51 +00:00
Ben Shi
fa2d31e9e6 [AVR] Fix a potential assert failure
Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D119416
2022-02-11 02:25:58 +00:00
Ayke van Laethem
316664783d
[AVR] Fix atomicrmw result value
This patch fixes the atomicrmw result value to be the value before the
operation instead of the value after the operation. This was a bug, left
as a FIXME in the code (see https://reviews.llvm.org/D97127).

From the LangRef:

> The contents of memory at the location specified by the <pointer>
> operand are atomically read, modified, and written back. The original
> value at the location is returned.

Doing this expansion early allows the register allocator to arrange
registers in such a way that commutable operations are simply swapped
around as needed, which results in shorter code while still being
correct.

Differential Revision: https://reviews.llvm.org/D117725
2022-02-02 09:10:39 +01:00
Benjamin Kramer
f15014ff54 Revert "Rename llvm::array_lengthof into llvm::size to match std::size from C++17"
This reverts commit ef8206320769ad31422a803a0d6de6077fd231d2.

- It conflicts with the existing llvm::size in STLExtras, which will now
  never be called.
- Calling it without llvm:: breaks C++17 compat
2022-01-26 16:55:53 +01:00
serge-sans-paille
ef82063207 Rename llvm::array_lengthof into llvm::size to match std::size from C++17
As a conquence move llvm::array_lengthof from STLExtras.h to
STLForwardCompat.h (which is included by STLExtras.h so no build
breakage expected).
2022-01-26 16:17:45 +01:00
Ayke van Laethem
116ab78694
[AVR] Make use of the constant value 0 in R1
The register R1 is defined to have the constant value 0 in the avr-gcc
calling convention (which we follow). Unfortunately, we don't really
make use of it. This patch replaces `LDI 0` instructions with a copy
from R1.

This reduces code size: my AVR build of compiler-rt goes from 50660 to
50240 bytes of code size, which is a 0.8% reduction. Presumably it will
also improve execution speed, although I didn't measure this.

Differential Revision: https://reviews.llvm.org/D117425
2022-01-23 17:08:01 +01:00
Jim Lin
d6b0734837 [NFC] Use Register instead of unsigned 2022-01-19 20:17:04 +08:00
Ben Shi
99e7bf46c9 [AVR] Optimize int16 shift operation for shift amount greater than 8
Skip operation on the lower byte in int16 logical left shift when
shift amount is greater than 8.

Skip operation on the higher byte in int16 logical & arithmetic
right shift when shift amount is greater than 8.

Reviewed By: aykevl

Differential Revision: https://reviews.llvm.org/D115594
2022-01-04 11:48:50 +00:00
Ben Shi
f4ef79306c [AVR] Optimize int8 arithmetic right shift 6 bits
Reviewed By: aykevl

Differential Revision: https://reviews.llvm.org/D115593
2022-01-04 10:36:03 +00:00
Ben Shi
9fb4e79d06 Revert "[AVR] Optimize int8 arithmetic right shift 6 bits"
This reverts commit 5723261370b45fa4d0d295845c6ef9e223f2ff4a.

There are failures as reported in

https://lab.llvm.org/buildbot#builders/16/builds/21638
https://lab.llvm.org/buildbot#builders/104/builds/5394
2022-01-04 04:14:15 +00:00
Ben Shi
5723261370 [AVR] Optimize int8 arithmetic right shift 6 bits
Reviewed By: aykevl

Differential Revision: https://reviews.llvm.org/D115593
2022-01-04 03:20:29 +00:00
Kazu Hirata
69ccc96162 [llvm] Use the default constructor for SDValue (NFC) 2022-01-01 10:36:59 -08:00
Kazu Hirata
5a667c0e74 [llvm] Use nullptr instead of 0 (NFC)
Identified with modernize-use-nullptr.
2021-12-28 08:52:25 -08:00
Zarko Todorovski
5b8bbbecfa [NFC][llvm] Inclusive language: reword and remove uses of sanity in llvm/lib/Target
Reworded removed code comments that contain `sanity check` and `sanity
test`.
2021-11-17 21:59:00 -05:00
Shivam Gupta
5449d2da65 [NFC] Run clang-format on llvm/lib/Trget/AVR/
The current inconsistency confuse contributors which coding guidlines to follow.
It would be better to have it consistent using clang-format tool.

Reviewed By: mhjacobson

Differential Revision: https://reviews.llvm.org/D109270
2021-09-04 20:05:15 +05:30
Arthur Eubanks
d7593ebaee [NFC] Clean up users of AttributeList::hasAttribute()
AttributeList::hasAttribute() is confusing, use clearer methods like
hasParamAttr()/hasRetAttr().

Add hasRetAttr() since it was missing from AttributeList.
2021-08-13 11:59:18 -07:00
Ayke van Laethem
4d7f5c0a85
[AVR] Only support sp, r0 and r1 in llvm.read_register
Most other registers are allocatable and therefore cannot be used.

This issue was flagged by the machine verifier, because reading other
registers is considered reading from an undefined register.

Differential Revision: https://reviews.llvm.org/D96969
2021-07-24 14:03:27 +02:00
Ayke van Laethem
feda08b70a
[AVR] Do not chain stores in call frame setup
Previously, AVRTargetLowering::LowerCall attempted to keep stack stores
in order with chains. Perhaps this worked in the past, but it does not
work now: it appears that the SelectionDAG legalization phase removes
these chains. Therefore, I've removed these chains entirely to match
X86 (which, similar to AVR, also prefers to use push instructions over
stack-relative stores to set up a call frame). With this change, all the
stack stores are in a somewhat reasonable order.

Differential Revision: https://reviews.llvm.org/D97853
2021-07-24 14:03:26 +02:00
Ben Shi
22668c6e1f [AVR][NFC] Refactor 8-bit & 16-bit shifts
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D98335
2021-05-31 10:30:46 +08:00
Ben Shi
86812faa5f [AVR] Improve inline assembly
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D96394
2021-05-30 23:44:43 +08:00
Dylan McKay
83e2710eb0 [AVR] Remove an assertion that causes generic CodeGen tests to fail
It was discussed a few years ago and agreed that it makes sense to
remove this assertion as other targets do not perform similar register
size checking in inline assembly constraint logic, so the check just
adds a needless barrier on AVR.

This patch removes the assertion and removes 'XFAIL' from two Generic
CodeGen tests for AVR as a result.
2021-02-05 02:05:23 +13:00
Kazu Hirata
8ed1636184 [llvm] Use isa instead of dyn_cast (NFC) 2021-01-29 23:23:37 -08:00
Ben Shi
50f1aa1db5 [AVR] Optimize 16-bit int shift
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D90092
2021-01-28 15:10:11 +08:00
Ben Shi
2a4acf3ea8 [AVR] Optimize 8-bit int shift
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D90678
2021-01-24 11:04:37 +08:00
Ben Shi
1eb8c5cd35 [AVR] Optimize 16-bit comparison with constant
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D93976
2021-01-24 00:38:57 +08:00
Ben Shi
25531a1d96 [AVR] Optimize 8-bit logic left/right shifts
Reviewed By: dylanmckay

Differential Revision: https://reviews.llvm.org/D89047
2021-01-23 23:54:16 +08:00
Ayke van Laethem
e03ba2198d
[AVR] Improve inline rotate/shift expansions
These expansions were rather inefficient and were done with more code
than necessary. This change optimizes them to use expansions more
similar to GCC. The code size is the same (when optimizing for code
size) but somehow LLVM reorders blocks in a non-optimal way. Still, this
should be an improvement with a reduction in code size of around 0.12%
(when building compiler-rt).

Differential Revision: https://reviews.llvm.org/D86418
2020-10-31 23:15:49 +01:00
Craig Topper
c193a689b4 [SelectionDAG] Use Align/MaybeAlign in calls to getLoad/getStore/getExtLoad/getTruncStore.
The versions that take 'unsigned' will be removed in the future.

I tried to use getOriginalAlign instead of getAlign in some
places. getAlign factors in the minimum alignment implied by
the offset in the pointer info. Since we're also passing the
pointer info we can use the original alignment.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D87592
2020-09-14 13:54:50 -07:00
Dylan McKay
b9c26a9cfe [AVR] Rewrite the function calling convention.
Summary:
The previous version relied on the standard calling convention using
std::reverse() to try to force the AVR ABI. But this only works for
simple cases, it fails for example with aggregate types.

This patch rewrites the calling convention with custom C++ code, that
implements the ABI defined in https://gcc.gnu.org/wiki/avr-gcc.

To do that it adds a few 16-bit pseudo registers for unaligned argument
passing, such as R24R23. For example this function:

    define void @fun({ i8, i16 } %a)

will pass %a.0 in R22 and %a.1 in R24R23.

There are no instructions that can use these pseudo registers, so a new
register class, DREGSMOVW, is defined to make them apart.

Also the ArgCC_AVR_BUILTIN_DIV is no longer necessary, as it is
identical to the C++ behavior (actually the clobber list is more strict
for __div* functions, but that is currently unimplemented).

Reviewers: dylanmckay

Subscribers: Gaelan, Sh4rK, indirect, jwagen, efriedma, dsprenkels, hiraditya, Jim, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D68524

Patch by Rodrigo Rivas Costa.
2020-06-23 21:36:18 +12:00
Guillaume Chatelet
1778564f91 [Alignment][NFC] Migrate the rest of backends
Summary: This is a followup on D81196

Reviewers: courbet

Subscribers: arsenm, dschuff, jyknight, dylanmckay, sdardis, nemanjai, jvesely, nhaehnle, sbc100, jgravelle-google, hiraditya, aheejin, kbarton, fedor.sergeev, asb, rbar, johnrusso, simoncook, sabuasal, niosHD, jrtc27, MaskRay, zzheng, edward-jones, atanasyan, rogfer01, MartinMosbeck, brucehoult, the_o, PkmX, jocewei, Jim, lenary, s.egerton, pzheng, sameer.abuasal, apazos, luismarques, kerbowa, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D81278
2020-06-08 07:17:20 +00:00
Ayke van Laethem
9505b5cb66
[AVR] Do not use divmod calls for bigger integers
The avr-libc provides *divmodqi4, *divmodhi4, and *divmodsi4 functions,
but does not provide a *divmoddi4. Instead it provides regular *divdi3
and *moddi3 functions.

Note that avr-libc doesn't support *divti3 or *modti3 for 128-bit
integer manipulation.

Source:
https://github.com/gcc-mirror/gcc/blob/releases/gcc-5.4.0/libgcc/config/avr/lib1funcs.S

Differential Revision: https://reviews.llvm.org/D78437
2020-04-20 13:56:38 +02:00
Dylan McKay
7b808b105f [AVR] Generalize the previous interrupt bugfix to signal handlers too 2020-03-31 19:33:34 +13:00
Dylan McKay
339b34266c [AVR] Respect the 'interrupt' function attribute
In the past, AVR functions were only lowered with interrupt-specific
machine code if the function was defined with the "avr-interrupt" or
"avr-signal" calling conventions.

This patch modifies the backend so that if the function does not have a
special calling convention, but does have an "interrupt" attribute,
that function is interpreted as a function with interrupts.

This also extracts the "is this function an interrupt" logic from
several disparate places in the backend into one AVRMachineFunctionInfo
attribute.

Bug found by Wilhelm Meier.
2020-03-31 19:00:18 +13:00
Simon Pilgrim
68224c1952 [TargetLowering] Only demand a rotation's modulo amount bits
ISD::ROTL/ROTR rotation values are guaranteed to act as a modulo amount, so for power-of-2 bitwidths we only need the lowest bits.

Differential Revision: https://reviews.llvm.org/D76201
2020-03-17 21:23:46 +00:00
Jim Lin
ea6eb813c7 [AVR][NFC] Use Register instead of unsigned
Summary: Use Register type for variables instead of unsigned type.

Reviewers: dylanmckay

Reviewed By: dylanmckay

Subscribers: hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D75595
2020-03-05 11:38:24 +08:00
Jim Lin
c0a2da9460 [AVR] Add missing ROLLOOP and RORLOOP into getTargetNodeName 2020-03-03 09:43:52 +08:00
Matt Arsenault
5fe4679cc9 AVR: Update for getRegisterByName change 2020-01-09 18:28:10 -05:00