On Mon, 2 Dec 2024 14:26:08 GMT, Maurizio Cimadamore
wrote:
>> As this is an internal API, I tried to keep it as similar as possible to the
>> other bulk operations. But as you say, the value used here for the last
>> parameter is always `segment.byteSize()`. We could easily change it. What's
On Mon, 2 Dec 2024 14:24:25 GMT, Maurizio Cimadamore
wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Improve docs
>
> test/jdk/java/foreign/TestStringEncoding.java line 263:
>
>> 261: for (var arena: a
On Mon, 2 Dec 2024 15:40:18 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no need
On Mon, 2 Dec 2024 10:14:54 GMT, Per Minborg wrote:
>> src/java.base/share/classes/jdk/internal/foreign/StringSupport.java line 128:
>>
>>> 126: */
>>> 127: @ForceInline
>>> 128: public static int strlenByte(final AbstractMemorySegmentImpl
>>> segment,
>>
>> Note: clients of this
On Mon, 2 Dec 2024 13:10:54 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no need
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
On Mon, 2 Dec 2024 13:56:59 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no need
On Mon, 2 Dec 2024 13:32:34 GMT, Martin Doerr wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Add zero detection method for quad words
>
> Thanks for the notification. The java/foreign suite has passed on linux on
On Mon, 2 Dec 2024 13:10:54 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no need
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
On Fri, 29 Nov 2024 12:44:53 GMT, Maurizio Cimadamore
wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Fix imports
>
> test/jdk/java/foreign/TestStringEncoding.java line 2:
>
>> 1: /*
>> 2: * Copyright (c) 2021, 2
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
On Fri, 29 Nov 2024 18:29:57 GMT, Jorn Vernee wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Fix imports
>
> src/java.base/share/classes/jdk/internal/foreign/SegmentBulkOperations.java
> line 205:
>
>> 203:
On Fri, 29 Nov 2024 12:43:28 GMT, Maurizio Cimadamore
wrote:
>> src/java.base/share/classes/jdk/internal/foreign/StringSupport.java line 187:
>>
>>> 185: // Prevent over scanning as we step by 2
>>> 186: final long endScan = toOffset & ~1; // The last bit is zero
>>> 187:
On Fri, 29 Nov 2024 12:19:52 GMT, Maurizio Cimadamore
wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Fix imports
>
> src/java.base/share/classes/jdk/internal/foreign/StringSupport.java line 128:
>
>> 126: */
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
On Fri, 29 Nov 2024 12:36:19 GMT, Maurizio Cimadamore
wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Fix imports
>
> src/java.base/share/classes/jdk/internal/foreign/StringSupport.java line 187:
>
>> 185:
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
On Fri, 29 Nov 2024 09:46:55 GMT, Per Minborg wrote:
>> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
>> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
>> 0`).
>>
>> This PR also improves performance on modern hardware, as there is no nee
On Fri, 29 Nov 2024 09:01:35 GMT, Maurizio Cimadamore
wrote:
>> Per Minborg has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Fix imports
>
> src/java.base/share/classes/jdk/internal/foreign/SegmentBulkOperations.java
> line 400:
>
>> 3
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods,
> fixing a bug in the `short_strlen` variant for odd offsets (`offset % 2 !=
> 0`).
>
> This PR also improves performance on modern hardware, as there is no need for
> pre-looping alignment. Removing this improves perfo
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods and
> move them to the class `SegmentBulkOperations` where other bulk operations
> reside.
>
> This PR fixes a bug in the `short_strlen` variant for offsets that were odd
> (`offset % 2 != 0`).
>
> This PR also improves
On Fri, 29 Nov 2024 07:55:25 GMT, Per Minborg wrote:
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods and
> move them to the class `SegmentBulkOperations` where other bulk operations
> reside.
>
> This PR fixes a bug in the `short_strlen` variant for offsets that were
On Fri, 29 Nov 2024 07:55:25 GMT, Per Minborg wrote:
> This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods and
> move them to the class `SegmentBulkOperations` where other bulk operations
> reside.
>
> This PR fixes a bug in the `short_strlen` variant for offsets that were
This PR proposes to rewrite the `StringSupport::chunkedStrlen*` methods and
move them to the class `SegmentBulkOperations` where other bulk operations
reside.
This PR fixes a bug in the `short_strlen` variant for offsets that were odd
(`offset % 2 != 0`).
This PR also improves performance on
30 matches
Mail list logo
