Appendix B: Cpu0 document and test

Cpu0 document

This section illustrates how to generate Cpu0 backend document.

Install sphinx

LLVM and this book use Sphinx to generate HTML documents. This book uses Sphinx to generate PDF and EPUB formats as well. See the installation guide here [1].

Sphinx uses reStructuredText format; see [8], [9], and [10]. For code-block formatting in this document, see [11] and [12].

On iMac you can install as follows:

brew install sphinx-doc
echo 'export PATH="/opt/homebrew/opt/sphinx-doc/bin:$PATH"' >> ~/.zshrc
% source ~/.zshrc

On Linux, install with:

sudo apt-get install python3-sphinx

The above installs Sphinx for making HTML documents but not PDF.

For PDF/LaTeX generation, install the following.

On iMac, install MacTeX.pkg from here [2] and restart your computer.

On Linux, install texlive as follows:

sudo apt-get install texlive texlive-latex-extra latexmk

or

sudo yum install texlive texlive-latex-extra latexmk

On Fedora 17, the texlive-latex-extra package is missing. Instead, install the package that includes pdflatex. For example, install pdfjam on Fedora 17 as follows,

[root@localhost lbd]$ yum list pdfjam
Loaded plugins: langpacks, presto, refresh-packagekit
Installed Packages
pdfjam.noarch                        2.08-3.fc17                         @fedora
[root@localhost lbd]$

Do

sudo apt-get update -y
sudo apt-get install -y latexmk

if latexmk gives error during ‘make latexpdf’ [3].

After upgrading to iMac OS X 10.11.1, the pdflatex link may be missing. Fix it by adding the following to your .profile:

114-37-153-62:lbd Jonathan$ ls /usr/local/texlive/2012/bin/universal-darwin/pdflatex
/usr/local/texlive/2012/bin/universal-darwin/pdflatex
114-37-153-62:lbd Jonathan$ cat ~/.profile
export PATH=$PATH:...:/usr/local/texlive/2012/bin/universal-darwin

Note of Sphinx

• Do not use more than 90 characters per line in code-block:: console or any other code block, otherwise errors may occur during make latexpdf.

• Use \clearpage (see gpu.rst) to force page breaks in pdf and avoid content splitting across two pages.

$HOME/git/lbd$ make latexpdf

Latexmk: applying rule 'pdflatex'...
Rule 'pdflatex': File changes, etc:
   Changed files, or newly in use since previous run(s):
  LLVMToolchainCpu0.aux
  LLVMToolchainCpu0.toc
Rule 'pdflatex': The following rules & subrules became out-of-date:
  pdflatex
Latexmk: Maximum runs of pdflatex reached without getting stable files
Latexmk: All targets (LLVMToolchainCpu0.pdf) are up-to-date
----------------------
This message may duplicate earlier message.
Latexmk: Failure in processing file 'LLVMToolchainCpu0.tex':
   'pdflatex' needed too many passes
----------------------
Latexmk: If appropriate, the -f option can be used to get latexmk
  to try to force complete processing.
make[1]: *** [LLVMToolchainCpu0.pdf] Error 12
make: *** [latexpdf] Error 2
make latexpdf  14.57s user 1.11s system 99% cpu 15.790 total

• Sphinx supports math symbols; see references [4] and [5] for more details.

Note of graphviz

• To draw nodes inside a parent node, use “compound=true;” and subgraph cluster_name [7]. To create edges from/to parent nodes, use the attributes “ltail” and “lhead” set to the parent node.

• See graphviz documents here [6].

Generate Cpu0 document

The Cpu0 example code is added chapter by chapter. It can be configured to a specific chapter by changing the CH definition in Cpu0SetChapter.h. For example, the following definition configures it to chapter 2.

lbdex/Cpu0/Cpu0SetChapter.h

#define CH       CH2

To help readers understand the backend structure step by step, the Cpu0 example code can be generated chapter by chapter using the following commands:

118-165-12-177:lbd Jonathan$ pwd
/home/Jonathan/test/lbd
118-165-12-177:lbd Jonathan$ make genexample
...
118-165-12-177:lbd Jonathan$ ls lbdex/chapters/
Chapter10_1  Chapter2    Chapter3_4  Chapter5_1  Chapter8_2
Chapter11_1  Chapter3_1  Chapter3_5  Chapter6_1  Chapter9_1
Chapter11_2  Chapter3_2  Chapter4_1  Chapter7_1  Chapter9_2
Chapter12_1  Chapter3_3  Chapter4_2  Chapter8_1  Chapter9_3

Besides the example code in each chapter, the above HTML and PDF Cpu0 documents also include the *.ll and *.s files located in the lbd/lbdex/output directory.

JonathantekiiMac:lbd Jonathan$ ls lbdex/output/
ch12_eh.cpu0.s                        ch12_thread_var.cpu0.pic.s      ch12_thread_var.ll
ch12_eh.ll                    ch12_thread_var.cpu0.static.s   ch4_math.s

Then, the HTML and PDF versions of this book can be generated by running the following commands.

118-165-12-177:lbd Jonathan$ pwd
/home/Jonathan/test/lbd
118-165-12-177:lbd Jonathan$ make html
...
118-165-12-177:lbd Jonathan$ make latexpdf
...

About Cpu0 Document

Since LLVM has a new release about every 6 months, and names of files, functions, classes, variables, etc. may change, maintaining the Cpu0 document is a continuous effort. The document adds code chapter by chapter.

To keep the document correct and easy to maintain, I use the :start-after: and :end-before: directives of reStructuredText to keep the document up to date.

For every new release, when the Cpu0 backend code changes, the document will reflect those changes in most of its content.

In the lbdex/Cpu0 folder, text beginning with //\@ and #ifdef CH > CHxx is referenced by document files *.rst.

In lbdex/llvm/modify/llvm, the *.rst files reference the code by copying it directly. Most references exist in llvmstructure.rst and elf.rst.

The example C/C++ code in lbdex/input comes from my own ideas and refers to the clang/test/CodeGen directory in the Clang source code release.

Cpu0 Regression Test

The last chapter can verify the Cpu0 backend’s generated code by using a Verilog simulator for code without global variable access.

The chapter lld in the repository https://github.com/Jonathan2251/lbt.git includes an LLVM ELF linker implementation and can verify test items involving global variable access.

However, LLVM provides its own test cases (regression tests) for each backend to verify the backend compiler [13] without needing a simulator or real hardware platform.

Cpu0 regression test items exist in the lbdex.tar.gz example code. Untar it into the lbdex/ directory.

For both iMac and Linux, copy the folder:

lbdex/regression-test/Cpu0

To

~/llvm/test/llvm/test/CodeGen/Cpu0

Then run the tests as follows on iMac, for both single and all test cases:

$ llvm-lit -a ~/llvm/test/llvm/test/CodeGen/Cpu0

The option -a shows the executing commands for each test case.

1-160-130-77:Cpu0 Jonathan$ pwd
/Users/Jonathan/llvm/test/llvm/test/CodeGen/Cpu0
1-160-130-77:Cpu0 Jonathan$ ~/llvm/test/build/bin/llvm-lit seteq.ll
-- Testing: 1 tests, 1 threads --
PASS: LLVM :: CodeGen/Cpu0/seteq.ll (1 of 1)
Testing Time: 0.08s
  Expected Passes    : 1
1-160-130-77:Cpu0 Jonathan$ ~/llvm/test/build/bin/llvm-lit .
...
PASS: LLVM :: CodeGen/Cpu0/zeroreg.ll
PASS: LLVM :: CodeGen/Cpu0/tailcall.ll
...

Run the following command to execute the test on Linux.

$ pwd
/home/cschen/llvm/test/llvm/test/CodeGen/Cpu0
$ ~/llvm/test/build/bin/llvm-lit seteq.ll
-- Testing: 1 tests, 1 threads --
PASS: LLVM :: CodeGen/Cpu0/seteq.ll (1 of 1)
Testing Time: 0.08s
  Expected Passes    : 1
$ ~/llvm/test/build/bin/llvm-lit .
...
PASS: LLVM :: CodeGen/Cpu0/zeroreg.ll
PASS: LLVM :: CodeGen/Cpu0/tailcall.ll
...

The chapters of this book and their related regression test items are listed as follows:

Table 52 Chapters

1	about
2	Cpu0 architecture and LLVM structure
3	Backend structure
4	Arithmetic and logic instructions
5	Generating object files
6	Global variables
7	Other data type
8	Control flow statements
9	Function call
10	ELF Support
11	Assembler
12	C++ support
13	Verify backend on verilog simulator

Table 53 Regression test items for Cpu0

File	v:pass x:fail	test ir, -> output asm	chapter
2008-06-05-Carry.ll	v		7
2008-07-15-InternalConstant.ll	v		6
2008-07-15-SmallSection.ll	v		6
2008-07-03-SRet.ll	v		9
2008-07-29-icmp.ll	v		8
2008-08-06-Alloca.ll	v		9
2008-08-01-AsmInline.ll	v		11
2008-08-08-ctlz.ll	v		7
2008-08-08-bswap.ll	v	bswap	12
2008-10-13-LegalizerBug.ll	v		8
2010-11-09-Mul.ll	v		4
2010-11-09-CountLeading.ll	v		7
2008-11-10-xint_to_fp.ll	v		7
addc.ll	v	64-bit add	7
addi.ll	v	32-bit add, sub	4
address-mode.ll	v	br, -> BB0_2:	8
alloca.ll	v	alloca i8, i32 %size, dynamic allocation	9
analyzebranch.ll	v	br, -> bne, beq	8
and1.ll	v	and	4
asm-large-immediate.ll	v	inline asm	11
atomic-1.ll	v	atomic	12
atomic-2.ll	v	atomic	12
atomics.ll	v	atomic	12
atomics-index.ll	v	atomic	12
atomics-fence.ll	v	atomic	12
br-jmp.ll	v	br, -> jmp	8
brockaddress.ll	v	blockaddress, -> lui, ori	8
cmov.ll	v	select, -> movn, movz	8
cprestore.ll	v	-> .cprestore	9
div.ll	v	sdiv, -> div, mflo	4
divrem.ll	v	sdiv, srem, udiv, urem, -> div, divu	4
div_rem.ll	v	sdiv, srem, -> div, mflo, mfhi	4
divu.ll	v	udiv, -> divu, mflo	4
divu_reml.ll	v	udiv, urem -> div, mflo, mfhi	4
double2int.ll	v	double to int, -> %call16(__fixdfsi)	7
eh-dwraf-cfa.ll	v		9
eh-return32.ll	v	Spill and reload all registers used for exception	9
eh.ll	v	c++ exception handling	12
ex2.ll	v	c++ exception handling	12
fastcc.ll	v	No effect in fastcc but can pass	9
fneg.ll	v	verify Cpu0 don’t uses hard float instruction	7
fp-spill-reload.ll	v	-> st $fp, ld $fp	9
frame-address.ll	v	addu $2, $zero, $fp	9
global-address.ll	v	global address, global variable	6
global-pointer.ll	v	global register load and retore, -> .cpload, .cprestore	9
gprestore.ll	v	global register retore, -> .cprestore	9
helloworld.ll	v	global register load and retore, -> .cpload, .cprestore	9
hf16_1.ll	v	function call in PIC, -> ld, jalr	9
i32k.ll	v	argument of constant int passing in register	9
i64arg.ll	v	argument of constant 64-bit passing in register	9
imm.ll	v	return constant 32-bit in register	9
indirectcall.ll	v	indirect function call	9
init-array.ll	v	check .init	6
inlineasm_constraint.ll	v	inline asm	11
inlineasm-cnstrnt-reg.ll	v		11
inlineasmmemop.ll	v		11
inlineasm-operand-code.ll	v		11
internalfunc.ll	v	internal function	9
jstat.ll	v	switch, -> JTI	8
lb1.ll	v	load i8*, sext i8, -> lb	7
lbu1.ll	v	load i8*, zext i8, -> lbu	7
lh1.ll	v	load i16*, sext i16, -> lh	7
lhu1.ll	v	load i16*, zext i16, -> lhu	7
llcarry.ll	v	64-bit add sub	7
longbranch.ll	v		8
machineverifier.ll	v	delay slot, (comment in machineverifier.ll)	8
mipslopat.ll	v	no check output (comment in mipslopat.ll)	6
misha.ll	v	miss alignment half word access	7
module-asm.ll	v	module asm	11
module-asm-cpu032II.ll	v	module asm	11
mul.ll	v	mul	4
mulll.ll	v	64-bit mul	4
mulull.ll	v	64-bit mul	4
not1.ll	v	not 1	4
null.ll	v	ret i32 0, -> ret $lr	3
o32_cc_byval.ll	v	by value	9
o32_cc_vararg.ll	v	variable argument	9
private.ll	v	private function call	9
rem.ll	v	srem, -> div, mfhi	4
remat-immed-load.ll	v	immediate load	3
remul.ll	v	urem, -> div, mfhi	4
return-vector-float4.ll	v	return vector, -> lui lui …	3
return-vector.ll	v	return vector, -> ld ld …, st st …	3
return_address.ll	v	llvm.returnaddress, -> addu $2, $zero, $lr	9
rotate.ll	v	rotl, rotr, -> rolv, rol, rorv	4
sb1.ll	v	store i8, sb	7
select.ll	v	select, -> movn, movz	8
seleq.ll	v	following for br with different condition	8
seleqk.ll	v		8
selgek.ll	v		8
selgt.ll	v		8
selle.ll	v		8
selltk.ll	v		8
selne.ll	v		8
selnek.ll	v		8
seteq.ll	v		8
seteqz.ll	v		8
setge.ll	v		8
setgek.ll	v		8
setle.ll	v		8
setlt.ll	v		8
setltk.ll	v		8
setne.ll	v		8
setuge.ll	v		8
setugt.ll	v		8
setule.ll	v		8
setult.ll	v		8
setultk.ll	v		8
sext_inreg.ll	v	sext i1, -> shl, sra	4
shift-parts.ll	v	64-bit shl, lshr, ashr, -> call function	9
shl1.ll	v	shl, -> shl	4
shl2.ll	v	shl, -> shlv	4
shr1.ll	v	shr, -> shr	4
shr2.ll	v	shr, -> shrv	4
sitofp-selectcc-opt.ll	v	comment in sitofp-selectcc-opt.ll	7
small-section-reserve-gp.ll	v	Cpu0 option -cpu0-use-small-section=true	6
sra1.ll	v	ashr, -> sra	4
sra2.ll	v	ashr, -> srav	4
stacksave-restore.ll	v		9
stacksize.ll	v	comment in stacksize.ll	9
stchar.ll	v	load and store i16, i8	7
stldst.ll	v	register sp spill	9
sub1.ll	v	sub, -> addiu	4
sub2.ll	v	sub, -> sub	4
tailcall.ll	v	tail call	9
tls.ll	v	ir thread_local global is for c++ “__thread int b;”	12
tls-alias.ll	v	thread_local global and thread local alias	12
tls-models.ll	v	ir external/internal thread_local global	12
uitofp.ll	v	integer2float, uitofp, -> jsub __floatunsisf	9
uli.ll	v	unalignment init, -> sb sb …	6
unalignedload.ll	v	unalignment init, -> sb sb …	6
vector-setcc.ll	v		7
weak.ll	v	extern_weak function, -> .weak	9
xor1.ll	v	xor, -> xor	4
zeroreg.ll	v	check register $zero	4

These supported test cases are located in lbdex/regression-test/Cpu0, which can be extracted from tar -xf lbdex.tar.gz.

The regression test is useful for two major reasons. First, it provides the LLVM input, assembly output, and the corresponding command and options within the same sample input file. This makes it a well-documented reference for both end users and developers.

Second, when developers make changes to their backend compiler—especially for optimization—these tests help detect side effects or bugs caused by the modifications. This is the core purpose of “regression testing.”

The following file includes the assembly output patterns for two subtargets of the Cpu0 backend. In addition to checking opcodes, it can also verify register numbers. For example, the destination register of the “andi” instruction must match the first source register of the following “xori” instruction. This is ensured when both are specified as register T1 in the corresponding assembly output.

lbdex/regression-test/Cpu0/setule.ll

; RUN: llc  -march=cpu0 -mcpu=cpu032I  -relocation-model=pic -O3 %s -o - | FileCheck %s -check-prefix=cpu032I
; RUN: llc  -march=cpu0 -mcpu=cpu032II  -relocation-model=pic -O3 %s -o - | FileCheck %s -check-prefix=cpu032II

@j = global i32 5, align 4
@k = global i32 10, align 4
@l = global i32 20, align 4
@m = global i32 10, align 4
@r1 = common global i32 0, align 4
@r2 = common global i32 0, align 4
@r3 = common global i32 0, align 4

define void @test() nounwind {
entry:
  %0 = load i32, i32* @j, align 4
  %1 = load i32, i32* @k, align 4
  %cmp = icmp ule i32 %0, %1
  %conv = zext i1 %cmp to i32
  store i32 %conv, i32* @r1, align 4
; cpu032I:  cmp	$sw, ${{[0-9]+|t9}}, ${{[0-9]+|t9}}
; cpu032I:  andi	$[[T1:[0-9]+|t9]], $sw, 1
; cpu032I:  xori	${{[0-9]+|t9}}, $[[T1]], 1
; cpu032II:  sltu	$[[T0:[0-9]+|t9]], ${{[0-9]+|t9}}, ${{[0-9]+|t9}}
; cpu032II:  xori	${{[0-9]+|t9}}, $[[T0]], 1
  %2 = load i32, i32* @m, align 4
  %cmp1 = icmp ule i32 %2, %1
  %conv2 = zext i1 %cmp1 to i32
  store i32 %conv2, i32* @r2, align 4
  ret void
}

Running regression tests must occur after building LLVM. However, the following README.rst and changes in related config files allow you to set up llvm-lit for a pre-built or pre-installed LLVM environment.

This setup enables running llvm-lit without rebuilding LLVM, making the regression testing more efficient when testing changes to the backend only.

lbdex/set-llvm-lit/README.txt

llvm-lit can be changed to support the installed clang/llvm. Then llvm-lit can
be run directly without build the clang/llvm.

## Run llvm-lit without building llvm-project

    $ pwd
    $ $HOME/test/llvm/llvm/test/CodeGen/Cpu0
    // change the following to dir for your llvm-project
    $ export LLVM_DIR=$HOME/test/llvm
    $ export LLVM_INSTALLED_DIR=$HOME/llvm-installed
    $ ~/llvm-installed/bin/llvm-lit addc.ll -a
    $ ~/llvm-installed/bin/llvm-lit .
    ~/test/llvm/clang/test/CodeGen/Cpu0$ ~/riscv/riscv_newlib/bin/llvm-lit . -a

set-llvm-lit % diff -r origin modify &> set-llvm-lit.diff

lbdex/set-llvm-lit/set-llvm-lit.diff

diff -r origin/bin/llvm-lit modify/bin/llvm-lit
8c8
< install_dir = r'/usr/local/riscv/andes/riscv_newlib'
---
> install_dir = os.environ["LLVM_INSTALLED_DIR"]
diff -r origin/test/Unit/lit.site.cfg.py modify/test/Unit/lit.site.cfg.py
3a4,9
> import os
> 
> llvm_project_dir = os.environ["PHOENIX_LLVM_DIR"]
> print('llvm_project_dir: ', llvm_project_dir)
> install_dir = os.environ["LLVM_INSTALLED_DIR"]
> 
4a11
> from pathlib import Path
7c14
<     return os.path.join(os.path.dirname(os.path.abspath(__file__)), p)
---
>     return str((Path(__file__).parent / p).resolve())
12,14c19,21
< config.llvm_src_root = path(r"/Users/cschen/llvm/test/llvm")
< config.llvm_obj_root = path(r"/Users/cschen/llvm/test/build")
< config.llvm_tools_dir = path(r"/Users/cschen/llvm/test/build/./bin")
---
> config.llvm_src_root = path(llvm_project_dir)+path(r"/llvm")
> config.llvm_obj_root = path(install_dir)
> config.llvm_tools_dir = path(install_dir)+path(r"/./bin")
17c24
< config.shlibdir = path(r"/Users/cschen/llvm/test/build/./lib")
---
> config.shlibdir = path(install_dir)+path(r"/./lib")
23a31
>     config.shlibdir = config.shlibdir % lit_config.params
diff -r origin/test/lit.site.cfg.py modify/test/lit.site.cfg.py
1c1
< # Autogenerated from /Users/cschen/llvm/test/llvm/test/lit.site.cfg.py.in
---
> # Autogenerated from /u/jonathanchen/andes/riscv/Phoenix/llvm-project/llvm/test/lit.site.cfg.py.in
3a4,9
> import os
> 
> llvm_project_dir = os.environ["PHOENIX_LLVM_DIR"]
> print('llvm_project_dir: ', llvm_project_dir)
> install_dir = r'/usr/local/riscv/andes/riscv_newlib'
> 
4a11
> from pathlib import Path
7c14
<     return os.path.join(os.path.dirname(os.path.abspath(__file__)), p)
---
>     return str((Path(__file__).parent / p).resolve())
12,19c19,27
< config.host_triple = "arm64-apple-darwin20.6.0"
< config.target_triple = "cpu0-unknown-elf"
< config.llvm_src_root = path(r"/Users/cschen/llvm/test/llvm")
< config.llvm_obj_root = path(r"/Users/cschen/llvm/test/build")
< config.llvm_tools_dir = path(r"/Users/cschen/llvm/test/build/./bin")
< config.llvm_lib_dir = path(r"/Users/cschen/llvm/test/build/./lib")
< config.llvm_shlib_dir = path(r"/Users/cschen/llvm/test/build/./lib")
< config.llvm_shlib_ext = ".dylib"
---
> config.host_triple = "x86_64-unknown-linux-gnu"
> config.target_triple = "riscv64-unknown-elf"
> config.llvm_src_root = path(llvm_project_dir)+path(r"/llvm")
> config.llvm_obj_root = path(install_dir)
> config.llvm_tools_dir = path(install_dir)+path(r"/./bin")
> config.llvm_lib_dir = path(install_dir)+path(r"/./lib")
> config.llvm_shlib_dir = path(install_dir)+path(r"/./lib")
> config.llvm_shlib_ext = ".so"
> config.llvm_plugin_ext = ".so"
22,24c30,34
< config.python_executable = "/opt/homebrew/Frameworks/Python.framework/Versions/3.9/bin/python3.9"
< config.gold_executable = "/opt/homebrew/bin/riscv64-unknown-elf-ld"
< config.ld64_executable = "/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ld"
---
> config.errc_messages = "No such file or directory;Is a directory;Invalid argument;Permission denied"
> config.python_executable = "/usr/bin/python3.6"
> config.gold_executable = "/usr/bin/ld.gold"
> config.ld64_executable = ""
> config.osx_sysroot = path(r"")
33,34c43,44
< config.targets_to_build = " Cpu0"
< config.native_target = "AArch64"
---
> config.targets_to_build = " RISCV"
> config.native_target = "X86"
36,38c46,48
< config.host_os = "Darwin"
< config.host_cc = "/opt/homebrew/opt/llvm/bin/clang "
< config.host_cxx = "/opt/homebrew/opt/llvm/bin/clang++ "
---
> config.host_os = "Linux"
> config.host_cc = "/usr/bin/cc "
> config.host_cxx = "/usr/bin/c++ "
44,45c54,55
< config.have_libxar = 1
< config.have_libxml2 = 1
---
> config.have_libxar = 0
> config.have_libxml2 = 0
52,53c62,63
< config.llvm_host_triple = 'arm64-apple-darwin20.6.0'
< config.host_arch = "arm64"
---
> config.llvm_host_triple = 'x86_64-unknown-linux-gnu'
> config.host_arch = "x86_64"
65a76
>     config.llvm_lib_dir = config.llvm_lib_dir % lit_config.params
diff -r origin/tools/clang/test/Unit/lit.site.cfg.py modify/tools/clang/test/Unit/lit.site.cfg.py
3a4,9
> import os
> 
> llvm_project_dir = os.environ["LLVM_DIR"]
> print('llvm_project_dir: ', llvm_project_dir)
> install_dir = os.environ["LLVM_INSTALLED_DIR"]
> 
4a11
> from pathlib import Path
7c14
<     return os.path.join(os.path.dirname(os.path.abspath(__file__)), p)
---
>     return str((Path(__file__).parent / p).resolve())
12,15c19,22
< config.llvm_src_root = path(r"/Users/cschen/llvm/test/llvm")
< config.llvm_obj_root = path(r"/Users/cschen/llvm/test/build")
< config.llvm_tools_dir = path(r"/Users/cschen/llvm/test/build/./bin")
< config.llvm_libs_dir = path(r"/Users/cschen/llvm/test/build/./lib")
---
> config.llvm_src_root = path(llvm_project_dir)+path(r"/llvm")
> config.llvm_obj_root = path(install_dir)
> config.llvm_tools_dir = path(install_dir)+path(r"/./bin")
> config.llvm_libs_dir = path(install_dir)+path(r"/./lib")
17c24
< config.clang_obj_root = path(r"/Users/cschen/llvm/test/build/tools/clang")
---
> config.clang_obj_root = path(install_dir)+path(r"/tools/clang")
19,20c26,27
< config.shlibdir = path(r"/Users/cschen/llvm/test/build/./lib")
< config.target_triple = "cpu0-unknown-elf"
---
> config.shlibdir = path(install_dir)+path(r"/./lib")
> config.target_triple = "riscv64-unknown-elf"
28a36
>     config.shlibdir = config.shlibdir % lit_config.params
36c44
<     config, os.path.join(path(r"/Users/cschen/llvm/test/clang"), "test/Unit/lit.cfg.py"))
---
>     config, os.path.join(path(llvm_project_dir)+path(r"/clang"), "test/Unit/lit.cfg.py"))
diff -r origin/tools/clang/test/lit.site.cfg.py modify/tools/clang/test/lit.site.cfg.py
3a4,9
> import os
> 
> llvm_project_dir = os.environ["LLVM_DIR"]
> print('llvm_project_dir: ', llvm_project_dir)
> install_dir = os.environ["LLVM_INSTALLED_DIR"]
> 
4a11
> from pathlib import Path
7c14
<     return os.path.join(os.path.dirname(os.path.abspath(__file__)), p)
---
>     return str((Path(__file__).parent / p).resolve())
12,17c19,24
< config.llvm_src_root = path(r"/Users/cschen/llvm/test/llvm")
< config.llvm_obj_root = path(r"/Users/cschen/llvm/test/build")
< config.llvm_tools_dir = path(r"/Users/cschen/llvm/test/build/./bin")
< config.llvm_libs_dir = path(r"/Users/cschen/llvm/test/build/./lib")
< config.llvm_shlib_dir = path(r"/Users/cschen/llvm/test/build/./lib")
< config.llvm_plugin_ext = ".dylib"
---
> config.llvm_src_root = path(llvm_project_dir)+path(r"/llvm")
> config.llvm_obj_root = path(install_dir)
> config.llvm_tools_dir = path(install_dir)+path(r"/./bin")
> config.llvm_libs_dir = path(install_dir)+path(r"/./lib")
> config.llvm_shlib_dir = path(install_dir)+path(r"/./lib")
> config.llvm_plugin_ext = ".so"
19,24c26,35
< config.clang_obj_root = path(r"/Users/cschen/llvm/test/build/tools/clang")
< config.clang_src_dir = path(r"/Users/cschen/llvm/test/clang")
< config.clang_tools_dir = path(r"/Users/cschen/llvm/test/build/./bin")
< config.host_triple = "arm64-apple-darwin20.6.0"
< config.target_triple = "cpu0-unknown-elf"
< config.host_cxx = "/opt/homebrew/opt/llvm/bin/clang++"
---
> config.errc_messages = "No such file or directory;Is a directory;Invalid argument;Permission denied"
> config.clang_lit_site_cfg = __file__
> config.clang_obj_root = path(install_dir)+path(r"/tools/clang")
> config.clang_src_dir = path(llvm_project_dir)+path(r"/clang")
> config.clang_tools_dir = path(install_dir)+path(r"/./bin")
> config.clang_lib_dir = path(install_dir)+path(r"/lib")
> config.host_triple = "x86_64-unknown-linux-gnu"
> config.target_triple = "riscv64-unknown-elf"
> config.host_cc = "/usr/bin/cc"
> config.host_cxx = "/usr/bin/c++"
34c45
< config.enable_experimental_new_pass_manager = 0
---
> config.enable_experimental_new_pass_manager = 1
36,37c47,48
< config.host_arch = "arm64"
< config.python_executable = "/opt/homebrew/Frameworks/Python.framework/Versions/3.9/bin/python3.9"
---
> config.host_arch = "x86_64"
> config.python_executable = "/usr/bin/python3.6"
40a52
> config.llvm_external_lit = path(r"")
diff -r origin/utils/lit/lit.site.cfg modify/utils/lit/lit.site.cfg
3a4,9
> import os
> 
> llvm_project_dir = os.environ["PHOENIX_LLVM_DIR"]
> print('llvm_project_dir: ', llvm_project_dir)
> install_dir = os.environ["LLVM_INSTALLED_DIR"]
> 
4a11
> from pathlib import Path
7c14
<     return os.path.join(os.path.dirname(os.path.abspath(__file__)), p)
---
>     return str((Path(__file__).parent / p).resolve())
13,15c20,22
< config.llvm_src_root = "/Users/cschen/llvm/test/llvm"
< config.llvm_obj_root = "/Users/cschen/llvm/test/build"
< config.llvm_tools_dir = "/Users/cschen/llvm/test/build/./bin"
---
> config.llvm_src_root = path(llvm_project_dir)+path(r"/llvm")
> config.llvm_obj_root = path(install_dir)
> config.llvm_tools_dir = path(install_dir)+path(r"/./bin")
30c37
< lit_config.load_config(config, "/Users/cschen/llvm/test/build/utils/lit/tests/lit.cfg")
---
> lit_config.load_config(config, os.path.join(config.llvm_obj_root, "utils/lit/tests/lit.cfg"))

• Only tools/clang/test/lit.site.cfg.py and test/lit.site.cfg.py need to be modified.

• The other files, tools/clang/test/Unit/lit.site.cfg.py, test/Unit/lit.site.cfg.py, and utils/lit/tests/lit.site.cfg.in, are empty and not used. However, I modified them as well for completeness.

[1]

https://www.sphinx-doc.org/en/master/usage/installation.html

[2]

http://www.tug.org/mactex/

[3]

https://zoomadmin.com/HowToInstall/UbuntuPackage/latexmk

[4]

https://sphinx-rtd-trial.readthedocs.io/en/latest/ext/math.html#module-sphinx.ext.mathbase

[5]

https://mirrors.mit.edu/CTAN/info/short-math-guide/short-math-guide.pdf

[6]

https://www.graphviz.org/pdf/dotguide.pdf

[7]

Ex. lbd/Fig/gpu/opengl-flow.gv. If the name of the subgraph begins with cluster, Graphviz notes the subgraph as a special cluster subgraph. If supported, the layout engine will do the layout so that the nodes belonging to the cluster are drawn together, with the entire drawing of the cluster contained within a bounding rectangle. https://graphviz.org/doc/info/lang.html.

[8]

http://docutils.sourceforge.net/docs/ref/rst/restructuredtext.html

[9]

http://docutils.sourceforge.net/docs/ref/rst/directives.html

[10]

http://docutils.sourceforge.net/rst.html

[11]

http://llvm.org/docs/SphinxQuickstartTemplate.html If you need to show LLVM IR use the llvm code block. https://llvm.org/docs/SphinxQuickstartTemplate.html#code-blocks

[12]

http://pygments.org/docs/lexers/

[13]

http://llvm.org/docs/TestingGuide.html