[lld] r228077 - Simplify large switches.
Rui Ueyama
ruiu at google.com
Wed Feb 4 18:08:37 PST 2015
Not ours, but some system .so files in Linux are actually linker scripts.
On Wed, Feb 4, 2015 at 6:04 PM, Sean Silva <chisophugis at gmail.com> wrote:
>
>
> On Wed, Feb 4, 2015 at 2:02 PM, Rui Ueyama <ruiu at google.com> wrote:
>
>> Comment is back in r228201.
>>
>> As to the performance, LLD took 200 *microseconds* in
>> GnuLdDriver::evaluateLinkerScript when linking itself on Linux (it parsed
>> 600 bytes of data). And these functions are only a small part of the entire
>> process. It's really negligible.
>>
>
> We use a linker script when self-hosting?
>
> -- Sean Silva
>
>
>>
>> On Tue, Feb 3, 2015 at 10:29 PM, Chandler Carruth <chandlerc at google.com>
>> wrote:
>>
>>> Comments are great, but I do much prefer the 'strchr' interface. So lets
>>> go with comments and a strchr interface.
>>>
>>> If this doesn't generate good code for some reason (no idea, haven't
>>> checked) we could either teach LLVM to generate awesome code for it or we
>>> could provide our own strchr-like utility that generates awesome code....
>>> But i'd like to see evidence that the performance of these is critical.
>>>
>>> On Tue, Feb 3, 2015 at 9:21 PM, Philip Reames <listmail at philipreames.com
>>> > wrote:
>>>
>>>> To me the first change is a readability regression (lost comments if
>>>> nothing else); the later two are neutral. I don't have a strong
>>>> preference, but please preserve the comments from the first function if
>>>> nothing else.
>>>>
>>>> p.s. Has anyone checked if we convert these strchrs to the "obvious"
>>>> bits of code? Each of these should be nearly as fast as the original
>>>> switches provided they're optimized correctly.
>>>>
>>>>
>>>> On 02/03/2015 06:33 PM, Shankar Easwaran wrote:
>>>>
>>>>> The style used in the old code is followed in clang too in various
>>>>> places. I felt its more readable too.
>>>>>
>>>>> This makes it less performant even if its not in performance critical
>>>>> pass and less readable too.
>>>>>
>>>>> Shankar Easwaran
>>>>>
>>>>> On 2/3/2015 6:00 PM, Rui Ueyama wrote:
>>>>>
>>>>>> Author: ruiu
>>>>>> Date: Tue Feb 3 18:00:21 2015
>>>>>> New Revision: 228077
>>>>>>
>>>>>> URL: http://llvm.org/viewvc/llvm-project?rev=228077&view=rev
>>>>>> Log:
>>>>>> Simplify large switches.
>>>>>>
>>>>>> This may be a little bit inefficient than the original code
>>>>>> but that should be okay as this is not really in a performance
>>>>>> critical pass.
>>>>>>
>>>>>> http://reviews.llvm.org/D7393
>>>>>>
>>>>>> Modified:
>>>>>> lld/trunk/lib/ReaderWriter/LinkerScript.cpp
>>>>>>
>>>>>> Modified: lld/trunk/lib/ReaderWriter/LinkerScript.cpp
>>>>>> URL: http://llvm.org/viewvc/llvm-project/lld/trunk/lib/
>>>>>> ReaderWriter/LinkerScript.cpp?rev=228077&r1=228076&r2=
>>>>>> 228077&view=diff
>>>>>> ==============================================================================
>>>>>>
>>>>>> --- lld/trunk/lib/ReaderWriter/LinkerScript.cpp (original)
>>>>>> +++ lld/trunk/lib/ReaderWriter/LinkerScript.cpp Tue Feb 3 18:00:21
>>>>>> 2015
>>>>>> @@ -242,66 +242,17 @@ bool Lexer::canStartNumber(char c) const
>>>>>> }
>>>>>> bool Lexer::canContinueNumber(char c) const {
>>>>>> - switch (c) {
>>>>>> - // Digits
>>>>>> - case '0': case '1': case '2': case '3': case '4': case '5': case
>>>>>> '6':
>>>>>> - case '7': case '8': case '9':
>>>>>> - case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
>>>>>> - case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
>>>>>> - // Hex marker
>>>>>> - case 'x': case 'X':
>>>>>> - // Type suffix
>>>>>> - case 'h': case 'H': case 'o': case 'O':
>>>>>> - // Scale suffix
>>>>>> - case 'M': case 'K':
>>>>>> - return true;
>>>>>> - default:
>>>>>> - return false;
>>>>>> - }
>>>>>> + return strchr("0123456789ABCDEFabcdefxXhHoOMK", c);
>>>>>> }
>>>>>> bool Lexer::canStartName(char c) const {
>>>>>> - switch (c) {
>>>>>> - case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case
>>>>>> 'G':
>>>>>> - case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case
>>>>>> 'N':
>>>>>> - case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case
>>>>>> 'U':
>>>>>> - case 'V': case 'W': case 'X': case 'Y': case 'Z':
>>>>>> - case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case
>>>>>> 'g':
>>>>>> - case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case
>>>>>> 'n':
>>>>>> - case 'o': case 'p': case 'q': case 'r': case 's': case 't': case
>>>>>> 'u':
>>>>>> - case 'v': case 'w': case 'x': case 'y': case 'z':
>>>>>> - case '_': case '.': case '$': case '/': case '\\':
>>>>>> - case '*':
>>>>>> - return true;
>>>>>> - default:
>>>>>> - return false;
>>>>>> - }
>>>>>> + return strchr(
>>>>>> + "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz_.$/\\*", c);
>>>>>> }
>>>>>> bool Lexer::canContinueName(char c) const {
>>>>>> - switch (c) {
>>>>>> - case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case
>>>>>> 'G':
>>>>>> - case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case
>>>>>> 'N':
>>>>>> - case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case
>>>>>> 'U':
>>>>>> - case 'V': case 'W': case 'X': case 'Y': case 'Z':
>>>>>> - case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case
>>>>>> 'g':
>>>>>> - case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case
>>>>>> 'n':
>>>>>> - case 'o': case 'p': case 'q': case 'r': case 's': case 't': case
>>>>>> 'u':
>>>>>> - case 'v': case 'w': case 'x': case 'y': case 'z':
>>>>>> - case '0': case '1': case '2': case '3': case '4': case '5': case
>>>>>> '6':
>>>>>> - case '7': case '8': case '9':
>>>>>> - case '_': case '.': case '$': case '/': case '\\': case '~': case
>>>>>> '=':
>>>>>> - case '+':
>>>>>> - case '[':
>>>>>> - case ']':
>>>>>> - case '*':
>>>>>> - case '?':
>>>>>> - case '-':
>>>>>> - case ':':
>>>>>> - return true;
>>>>>> - default:
>>>>>> - return false;
>>>>>> - }
>>>>>> + return strchr("ABCDEFGHIJKLMNOPQRSTUVWXYZabcd
>>>>>> efghijklmnopqrstuvwxyz"
>>>>>> + "0123456789_.$/\\~=+[]*?-:", c);
>>>>>> }
>>>>>> /// Helper function to split a StringRef in two at the nth
>>>>>> character.
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> llvm-commits mailing list
>>>>>> llvm-commits at cs.uiuc.edu
>>>>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>> _______________________________________________
>>>> llvm-commits mailing list
>>>> llvm-commits at cs.uiuc.edu
>>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>>>
>>>
>>>
>>> _______________________________________________
>>> llvm-commits mailing list
>>> llvm-commits at cs.uiuc.edu
>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>>
>>>
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150204/e879cb27/attachment.html>
More information about the llvm-commits
mailing list