Use computed-goto to lower instruction dispatch overhead #2
+115
−25
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jserv
reviewed
Dec 29, 2020
jserv
reviewed
Dec 29, 2020
jserv
reviewed
Dec 29, 2020
jserv
reviewed
Dec 29, 2020
jserv
reviewed
Dec 30, 2020
jserv
reviewed
Dec 30, 2020
jserv
requested changes
Dec 30, 2020
jserv
reviewed
Dec 30, 2020
|
The commit message should be amended, mentioning the benefit of the proposed computed-goto change. |
+ This commit applies the computed goto technique to the main loop. This modification is tested to have 15 to 23 percent improvement on performance. + The old implementation of rv_step can be enabled by setting the environment variable ENABLE_COMPUTED_GOTO to 0. + Only clang and gcc is supported by computed goto feature, ENABLE_COMPUTED_GOTO will be ignored if other compiler is used. + An op_unimp handler is also added in order to acheive this. This also allows better handling of unimplemented opcodes instead of jumping into NULL.
sammer1107
force-pushed
the
computed_goto
branch
from
7087c7f to
3ecac53
Compare
sammer1107
changed the title
jserv
reviewed
Jan 5, 2021
jserv
requested changes
Jan 5, 2021
jserv
reviewed
Jan 6, 2021
jserv
reviewed
Jan 9, 2021
|
@sammer1107, would you continue this work recently? |
|
Thank @sammer1107 for contributing! |
vacantron
pushed a commit
to vacantron/rv32emu
that referenced
this pull request
Dec 14, 2024
-fsanitize=address report when CTRL+a x exiting:
=================================================================
==297977==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 23400 byte(s) in 117 object(s) allocated from:
#0 0x761b706fd340 in calloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:77
#1 0x633ad9c5e310 in block_translate src/emulate.c:649
sysprog21#2 0x633ad9c5e310 in block_find_or_translate src/emulate.c:865
sysprog21#3 0x633ad9c5e310 in rv_step src/emulate.c:1029
sysprog21#4 0x633ad9c5e310 in rv_run src/riscv.c:498
sysprog21#5 0x633ad9c5e310 in main src/main.c:279
Direct leak of 3136 byte(s) in 125 object(s) allocated from:
#0 0x761b706fd9c7 in malloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:69
#1 0x633ad9c5fea4 in match_pattern src/emulate.c:767
sysprog21#2 0x633ad9c5fea4 in block_find_or_translate src/emulate.c:872
sysprog21#3 0x633ad9c5fea4 in rv_step src/emulate.c:1029
sysprog21#4 0x633ad9c5fea4 in rv_run src/riscv.c:498
sysprog21#5 0x633ad9c5fea4 in main src/main.c:279
Register a clean up callback, async_block_clear() to free all the
allocated memory fix this when emulator exits.
vestata
pushed a commit
to vestata/rv32emu
that referenced
this pull request
Jan 24, 2025
Use computed-goto to lower instruction dispatch overhead
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
main loop. This modification is tested to have 15 to 23
percent improvement on performance.
by setting the environment variable ENABLE_COMPUTED_GOTO to 0.
ENABLE_COMPUTED_GOTO will be ignored if other compiler is used.
this. This also allows better handling of unimplemented
opcodes instead of jumping into NULL.