diff options
| author | Christian Krinitsin <mail@krinitsin.com> | 2025-07-17 09:10:43 +0200 |
|---|---|---|
| committer | Christian Krinitsin <mail@krinitsin.com> | 2025-07-17 09:10:43 +0200 |
| commit | f2ec263023649e596c5076df32c2d328bc9393d2 (patch) | |
| tree | 5dd86caab46e552bd2e62bf9c4fb1a7504a44db4 /results/scraper/fex/1732 | |
| parent | 63d2e9d409831aa8582787234cae4741847504b7 (diff) | |
| download | qemu-analysis-main.tar.gz qemu-analysis-main.zip | |
Diffstat (limited to 'results/scraper/fex/1732')
| -rw-r--r-- | results/scraper/fex/1732 | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/results/scraper/fex/1732 b/results/scraper/fex/1732 new file mode 100644 index 000000000..655145298 --- /dev/null +++ b/results/scraper/fex/1732 @@ -0,0 +1,8 @@ +TME extension: ideas on how to (ab)use it +Following up from yesterday's discussion, + +For `tso`/`non-tso` memory tracking, we might be able to group accesses between atomic ops in transactions, and detect concurrent access with transaction aborts. Might get us more precise tracking than the [MTE](https://github.com/FEX-Emu/FEX/issues/1731) approach. + +For synchronous detection of segfaults, we might be able to wrap code between guest abi sync points around a transaction, and on transaction abort execute some slowpath. Details on synchronous signal complications in #1682 + +TME Extension details: https://developer.arm.com/documentation/101028/0012/16--Transactional-Memory-Extension--TME--intrinsics?lang=en \ No newline at end of file |