RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-22 10:42:39 +01:00
Code Issues Projects Releases Wiki Activity

Speedup some unicode rendering

Browse Source 
Use a fast path for column width computation for ascii characters. Especially
relevant for llvm-objdump.

before:

    % time ./bin/llvm-objdump -D  -j .text /lib/libc.so.6 >/dev/null
    ./bin/llvm-objdump -D -j .text /lib/libc.so.6 > /dev/null  0.75s user 0.01s system 99% cpu 0.757 total

after:

    % time ./bin/llvm-objdump -D  -j .text /lib/libc.so.6 >/dev/null
    ./bin/llvm-objdump -D -j .text /lib/libc.so.6 > /dev/null  0.37s user 0.01s system 99% cpu 0.378 total

Differential Revision: https://reviews.llvm.org/D92180
This commit is contained in:
serge-sans-paille 2020-11-26 15:22:20 +01:00
parent 5d70260f30
commit 8db0e1abee
2 changed files with 26 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
lib/Support/Unicode.cpp
View File

@ -339,11 +339,22 @@ static inline int charWidth(int UCS)
return 1;
}
static bool isprintableascii(char c) { return c > 31 && c < 127; }
int columnWidthUTF8(StringRef Text) {
unsigned ColumnWidth = 0;
unsigned Length;
for (size_t i = 0, e = Text.size(); i < e; i += Length) {
Length = getNumBytesForUTF8(Text[i]);
// fast path for ASCII characters
if (Length == 1) {
if (!isprintableascii(Text[i]))
return ErrorNonPrintableCharacter;
ColumnWidth += 1;
continue;
}
if (Length <= 0 || i + Length > Text.size())
return ErrorInvalidUTF8;
UTF32 buf[1];

15
unittests/Support/UnicodeTest.cpp
View File

@ -7,6 +7,7 @@
//===----------------------------------------------------------------------===//
#include "llvm/Support/Unicode.h"
#include "llvm/Support/ConvertUTF.h"
#include "gtest/gtest.h"
namespace llvm {
@ -23,6 +24,7 @@ TEST(Unicode, columnWidthUTF8) {
EXPECT_EQ(6, columnWidthUTF8("abcdef"));
EXPECT_EQ(-1, columnWidthUTF8("\x01"));
EXPECT_EQ(-1, columnWidthUTF8("\t"));
EXPECT_EQ(-1, columnWidthUTF8("aaaaaaaaaa\x01"));
EXPECT_EQ(-1, columnWidthUTF8("\342\200\213")); // 200B ZERO WIDTH SPACE
@ -84,6 +86,19 @@ TEST(Unicode, isPrintable) {
EXPECT_TRUE(isPrintable(0x20000)); // CJK UNIFIED IDEOGRAPH-20000
EXPECT_FALSE(isPrintable(0x10FFFF)); // noncharacter
// test the validity of a fast path in columnWidthUTF8
for (unsigned char c = 0; c < 128; ++c) {
const UTF8 buf8[2] = {c, 0};
const UTF8 *Target8 = &buf8[0];
UTF32 buf32[1];
UTF32 *Target32 = &buf32[0];
auto status = ConvertUTF8toUTF32(&Target8, Target8 + 1, &Target32,
Target32 + 1, strictConversion);
EXPECT_TRUE(status == conversionOK);
EXPECT_TRUE((columnWidthUTF8(reinterpret_cast<const char *>(buf8)) == 1) ==
(bool)isPrintable(buf32[0]));
}
}
} // namespace