[Patch] allows new memory info name for WebGPU#27475
Merged
tianleiwu merged 1 commit intotlwu/rel-1.24.3from Feb 27, 2026
Merged
[Patch] allows new memory info name for WebGPU#27475tianleiwu merged 1 commit intotlwu/rel-1.24.3from
tianleiwu merged 1 commit intotlwu/rel-1.24.3from
Conversation
Contributor
There was a problem hiding this comment.
Pull request overview
Extends the C API OrtApis::CreateMemoryInfo handling in ORT 1.24.x to accept the newer shortened WebGPU (and WebNN) memory info names used by newer (1.25.x) plugin EP builds, improving cross-version plugin compatibility.
Changes:
- Accept
"WebGPU_Buf"as an alias for"WebGPU_Buffer"inCreateMemoryInfo. - Accept
"WebNN_Ten"as an alias for"WebNN_Tensor"inCreateMemoryInfo. - Add inline comments explaining the origin of the shortened names (SSO on wasm32) and the compatibility intent.
Comments suppressed due to low confidence (1)
onnxruntime/core/framework/allocator.cc:256
- CreateMemoryInfo now accepts the short aliases ("WebGPU_Buf"/"WebNN_Ten"), but it also stores
name1verbatim into OrtMemoryInfo. In 1.24.x many call sites compareOrtMemoryInfo::nameto the canonical constants (e.g.,onnxruntime::WEBGPU_BUFFER/WEBNN_TENSOR), andOrtMemoryInfoequality/hashing includesname, so passing the short alias can lead to allocator lookup / device checks failing later. Consider normalizing these aliases to the existing canonical names when constructingOrtMemoryInfo, or otherwise ensure all downstream comparisons treat both names as equivalent.
} else if (strcmp(name1, onnxruntime::WEBGPU_BUFFER) == 0 ||
strcmp(name1, onnxruntime::WEBNN_TENSOR) == 0 ||
// PR #27207 (merged to main/1.25.x, not in 1.24.x) shortened the WebGPU/WebNN
// memory info names from "WebGPU_Buffer"/"WebNN_Tensor" to "WebGPU_Buf"/"WebNN_Ten"
// to enable Small String Optimization (SSO) on wasm32 (emscripten), where strings
// must be <= 10 chars for SSO.
//
// A WebGPU/WebNN plugin EP built against 1.25.x will use the new short names.
// Accept both old and new names here so that plugin EPs targeting either 1.24.x
// or 1.25.x can work with this 1.24.x runtime.
//
// See: https://github.com/microsoft/onnxruntime/pull/27207
strcmp(name1, "WebGPU_Buf") == 0 ||
strcmp(name1, "WebNN_Ten") == 0) {
*out = new OrtMemoryInfo(
name1, type,
OrtDevice(OrtDevice::GPU, OrtDevice::MemType::DEFAULT, OrtDevice::VendorIds::NONE, device_id),
mem_type1);
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+240
to
+252
| strcmp(name1, onnxruntime::WEBNN_TENSOR) == 0 || | ||
| // PR #27207 (merged to main/1.25.x, not in 1.24.x) shortened the WebGPU/WebNN | ||
| // memory info names from "WebGPU_Buffer"/"WebNN_Tensor" to "WebGPU_Buf"/"WebNN_Ten" | ||
| // to enable Small String Optimization (SSO) on wasm32 (emscripten), where strings | ||
| // must be <= 10 chars for SSO. | ||
| // | ||
| // A WebGPU/WebNN plugin EP built against 1.25.x will use the new short names. | ||
| // Accept both old and new names here so that plugin EPs targeting either 1.24.x | ||
| // or 1.25.x can work with this 1.24.x runtime. | ||
| // | ||
| // See: https://github.com/microsoft/onnxruntime/pull/27207 | ||
| strcmp(name1, "WebGPU_Buf") == 0 || | ||
| strcmp(name1, "WebNN_Ten") == 0) { |
There was a problem hiding this comment.
This change also adds support for the WebNN short name ("WebNN_Ten"), but the PR title/description only mention WebGPU. Please update the PR description (or remove the WebNN part) so the change scope matches what’s being shipped.
Comment on lines
+241
to
+252
| // PR #27207 (merged to main/1.25.x, not in 1.24.x) shortened the WebGPU/WebNN | ||
| // memory info names from "WebGPU_Buffer"/"WebNN_Tensor" to "WebGPU_Buf"/"WebNN_Ten" | ||
| // to enable Small String Optimization (SSO) on wasm32 (emscripten), where strings | ||
| // must be <= 10 chars for SSO. | ||
| // | ||
| // A WebGPU/WebNN plugin EP built against 1.25.x will use the new short names. | ||
| // Accept both old and new names here so that plugin EPs targeting either 1.24.x | ||
| // or 1.25.x can work with this 1.24.x runtime. | ||
| // | ||
| // See: https://github.com/microsoft/onnxruntime/pull/27207 | ||
| strcmp(name1, "WebGPU_Buf") == 0 || | ||
| strcmp(name1, "WebNN_Ten") == 0) { |
There was a problem hiding this comment.
There’s no regression test ensuring CreateMemoryInfo accepts the new short aliases and produces a usable memory info for subsequent allocator/device matching. Adding a small shared_lib test that calls the public C/C++ API with "WebGPU_Buf" (and optionally "WebNN_Ten") would help prevent future breakage of this compatibility behavior.
This was referenced Mar 9, 2026
fs-eire
added a commit
that referenced
this pull request
Mar 12, 2026
Accept pre-1.25 names "WebGPU_Buffer"/"WebNN_Tensor" as aliases in CreateMemoryInfo and normalize them to the current short names "WebGPU_Buf"/"WebNN_Ten". This allows released onnxruntime-genai (which still uses the old long names) to work with main branch ORT. The normalization is critical because downstream code (external data loader, WebGPU context) compares memory info names against the current constants. Without normalization, those comparisons would fail. See: #27207 See: #27475 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
fs-eire
added a commit
that referenced
this pull request
Mar 13, 2026
…ty (#27637) ### Description Accept pre-1.25 names `"WebGPU_Buffer"`/`"WebNN_Tensor"` as aliases in `CreateMemoryInfo` and normalize them to the current short names `"WebGPU_Buf"`/`"WebNN_Ten"`. This is the **reverse** of #27475 (which added forward compatibility in the 1.24.x patch branch). ### Motivation and Context Released onnxruntime-genai still uses the old (pre-1.25) long names when calling `CreateMemoryInfo`. Without this change, those calls fail with `ORT_INVALID_ARGUMENT` on main branch. ### Key Design Decision When an old name is detected, it is **normalized** to the current short constant (e.g., `"WebGPU_Buffer"` -> `"WebGPU_Buf"`). This is critical because downstream code (e.g., `external_data_loader.cc`, `webgpu_context.cc`) compares `OrtMemoryInfo.name` against the current constants. Simply passing through the old name would cause those comparisons to fail. ### Changes - `onnxruntime/core/framework/allocator.cc`: Accept and normalize legacy names in `CreateMemoryInfo` - `onnxruntime/test/shared_lib/test_allocator.cc`: Add test verifying legacy names are accepted and normalized ### See Also - #27207 (original rename) - #27475 (forward compat in 1.24.x) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This was referenced Mar 16, 2026
deps(nuget): Bump the microsoft-packages group with 8 updates
Ellerbach/azure-ai-search-simulator#76
Closed
This was referenced Mar 23, 2026
deps(nuget): Bump the microsoft-packages group with 8 updates
Ellerbach/azure-ai-search-simulator#80
Closed
This was referenced Mar 30, 2026
This file contains hidden or 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
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.
Description
allows new memory info name for WebGPU.
Motivation and Context
This allows at least 1.24.3 works with future (1.25.x) WebGPU plugin DLL