From 4c2d6332a97db6e51bbe39735dfbd659246fb80b Mon Sep 17 00:00:00 2001
From: Gerrit User 1571352 <1571352@3ce6091f-6c88-37e8-8c75-72f92ae8dfba>
Date: Thu, 12 Dec 2024 02:58:31 -0800
Subject: [PATCH] Update patch set 3

Patch Set 3:

(2 comments)

Patch-set: 3
Attention: {"person_ident":"Gerrit User 1571352 \u003c1571352@3ce6091f-6c88-37e8-8c75-72f92ae8dfba\u003e","operation":"REMOVE","reason":"\u003cGERRIT_ACCOUNT_1571352\u003e replied on the change"}
Attention: {"person_ident":"Gerrit User 1115898 \u003c1115898@3ce6091f-6c88-37e8-8c75-72f92ae8dfba\u003e","operation":"ADD","reason":"\u003cGERRIT_ACCOUNT_1571352\u003e replied on the change"}
---
 d5818026ec4bd861a5a8aba6d428c576ba2ba5f5 | 18 ++++++++++++++++++
 f855222902c6570dfb3dd0ec23c65dc55bc326c8 | 21 +++++++++++++++++++++
 2 files changed, 39 insertions(+)
 create mode 100644 f855222902c6570dfb3dd0ec23c65dc55bc326c8

diff --git a/d5818026ec4bd861a5a8aba6d428c576ba2ba5f5 b/d5818026ec4bd861a5a8aba6d428c576ba2ba5f5
index b8b7a999e..bbef7bdd5 100644
--- a/d5818026ec4bd861a5a8aba6d428c576ba2ba5f5
+++ b/d5818026ec4bd861a5a8aba6d428c576ba2ba5f5
@@ -16,6 +16,24 @@
       "message": "ideally run sme benchmarks on apple m4 or equivalent.\nconsider doing LDP with int64 to achieve 128 bit load/store",
       "revId": "d5818026ec4bd861a5a8aba6d428c576ba2ba5f5",
       "serverId": "3ce6091f-6c88-37e8-8c75-72f92ae8dfba"
+    },
+    {
+      "unresolved": false,
+      "key": {
+        "uuid": "aae285fe_0f0426f4",
+        "filename": "/PATCHSET_LEVEL",
+        "patchSetId": 2
+      },
+      "lineNbr": 0,
+      "author": {
+        "id": 1571352
+      },
+      "writtenOn": "2024-12-12T10:58:31Z",
+      "side": 1,
+      "message": "SVE and the first version of SME do not have an LDP equivalent instruction. SVE 2.1 and SME 2 introduce multi-vector load/store instructions which are equivalent to LDP/STP however these would need additional compiler features to be enabled (-march\u003d...+sme2 rather than -march\u003d...+sme which we currently use).\n\nOf course if you can guarantee that the vector length is 128 bits then you can use the Neon LDP/STP instructions since they share the same vector length, but this is not portable so I would prefer to avoid hard-coding this anywhere.",
+      "parentUuid": "6f23531a_fee498dc",
+      "revId": "d5818026ec4bd861a5a8aba6d428c576ba2ba5f5",
+      "serverId": "3ce6091f-6c88-37e8-8c75-72f92ae8dfba"
     }
   ]
 }
\ No newline at end of file
diff --git a/f855222902c6570dfb3dd0ec23c65dc55bc326c8 b/f855222902c6570dfb3dd0ec23c65dc55bc326c8
new file mode 100644
index 000000000..dfe726783
--- /dev/null
+++ b/f855222902c6570dfb3dd0ec23c65dc55bc326c8
@@ -0,0 +1,21 @@
+{
+  "comments": [
+    {
+      "unresolved": false,
+      "key": {
+        "uuid": "d7c4ab9b_462f39a0",
+        "filename": "/PATCHSET_LEVEL",
+        "patchSetId": 3
+      },
+      "lineNbr": 0,
+      "author": {
+        "id": 1571352
+      },
+      "writtenOn": "2024-12-12T10:58:31Z",
+      "side": 1,
+      "message": "(rebased)",
+      "revId": "f855222902c6570dfb3dd0ec23c65dc55bc326c8",
+      "serverId": "3ce6091f-6c88-37e8-8c75-72f92ae8dfba"
+    }
+  ]
+}
\ No newline at end of file