Skip to content

metal: log recommendedMaxWorkingSetSize on iOS 16+ #4936

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Jan 16, 2024
Merged

metal: log recommendedMaxWorkingSetSize on iOS 16+ #4936

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
e26d224
metal: Log `recommendedMaxWorkingSetSize` on iOS 16+
azarovalex Jan 14, 2024
3c4d2a1
Only log on iOS and macOS, ignoring tvOS and other platforms
azarovalex Jan 14, 2024
ed36ef8
Check for Xcode version before using recommendedMaxWorkingSetSize
azarovalex Jan 14, 2024
d84f5c2
Merge branch 'master' into azarovalex/recommendedMaxWorkingSetSize
ggerganov Jan 16, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
58 changes: 27 additions & 31 deletions ggml-metal.m
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -369,8 +369,12 @@ static void ggml_metal_log(enum ggml_log_level level, const char * format, ...){
GGML_METAL_LOG_INFO("%s: simdgroup reduction support = %s\n", __func__, ctx->support_simdgroup_reduction ? "true" : "false");
GGML_METAL_LOG_INFO("%s: simdgroup matrix mul. support = %s\n", __func__, ctx->support_simdgroup_mm ? "true" : "false");
GGML_METAL_LOG_INFO("%s: hasUnifiedMemory = %s\n", __func__, ctx->device.hasUnifiedMemory ? "true" : "false");
#if TARGET_OS_OSX
GGML_METAL_LOG_INFO("%s: recommendedMaxWorkingSetSize = %8.2f MB\n", __func__, ctx->device.recommendedMaxWorkingSetSize / 1e6);

#if TARGET_OS_OSX || (TARGET_OS_IOS && __clang_major__ >= 15)
if (@available(macOS 10.12, iOS 16.0, *)) {
GGML_METAL_LOG_INFO("%s: recommendedMaxWorkingSetSize = %8.2f MB\n", __func__, ctx->device.recommendedMaxWorkingSetSize / 1e6);
}
#elif TARGET_OS_OSX
if (ctx->device.maxTransferRate != 0) {
GGML_METAL_LOG_INFO("%s: maxTransferRate = %8.2f MB/s\n", __func__, ctx->device.maxTransferRate / 1e6);
} else {
Expand Down Expand Up @@ -2369,6 +2373,25 @@ GGML_CALL static void ggml_backend_metal_buffer_clear(ggml_backend_buffer_t buff
UNUSED(buft);
}

static void ggml_backend_metal_log_allocated_size(id<MTLDevice> device) {
#if TARGET_OS_OSX || (TARGET_OS_IOS && __clang_major__ >= 15)
if (@available(macOS 10.12, iOS 16.0, *)) {
GGML_METAL_LOG_INFO(", (%8.2f / %8.2f)",
device.currentAllocatedSize / 1024.0 / 1024.0,
device.recommendedMaxWorkingSetSize / 1024.0 / 1024.0);

if (device.currentAllocatedSize > device.recommendedMaxWorkingSetSize) {
GGML_METAL_LOG_WARN("%s: warning: current allocated size is greater than the recommended max working set size\n", __func__);
} else {
GGML_METAL_LOG_INFO("\n");
}
} else {
GGML_METAL_LOG_INFO(", (%8.2f)\n", device.currentAllocatedSize / 1024.0 / 1024.0);
}
#endif
UNUSED(device);
}

GGML_CALL static ggml_backend_buffer_t ggml_backend_metal_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft, size_t size) {
struct ggml_backend_metal_buffer_context * ctx = malloc(sizeof(struct ggml_backend_metal_buffer_context));

Expand Down Expand Up @@ -2401,22 +2424,7 @@ GGML_CALL static ggml_backend_buffer_t ggml_backend_metal_buffer_type_alloc_buff
}

GGML_METAL_LOG_INFO("%s: allocated buffer, size = %8.2f MiB", __func__, size_aligned / 1024.0 / 1024.0);


#if TARGET_OS_OSX
GGML_METAL_LOG_INFO(", (%8.2f / %8.2f)",
device.currentAllocatedSize / 1024.0 / 1024.0,
device.recommendedMaxWorkingSetSize / 1024.0 / 1024.0);

if (device.currentAllocatedSize > device.recommendedMaxWorkingSetSize) {
GGML_METAL_LOG_WARN("%s: warning: current allocated size is greater than the recommended max working set size\n", __func__);
} else {
GGML_METAL_LOG_INFO("\n");
}
#else
GGML_METAL_LOG_INFO(", (%8.2f)\n", device.currentAllocatedSize / 1024.0 / 1024.0);
#endif

ggml_backend_metal_log_allocated_size(device);

return ggml_backend_buffer_init(buft, ggml_backend_metal_buffer_i, ctx, size);
}
Expand Down Expand Up @@ -2524,19 +2532,7 @@ GGML_CALL ggml_backend_buffer_t ggml_backend_metal_buffer_from_ptr(void * data,
}
}

#if TARGET_OS_OSX
GGML_METAL_LOG_INFO(", (%8.2f / %8.2f)",
device.currentAllocatedSize / 1024.0 / 1024.0,
device.recommendedMaxWorkingSetSize / 1024.0 / 1024.0);

if (device.currentAllocatedSize > device.recommendedMaxWorkingSetSize) {
GGML_METAL_LOG_WARN("%s: warning: current allocated size is greater than the recommended max working set size\n", __func__);
} else {
GGML_METAL_LOG_INFO("\n");
}
#else
GGML_METAL_LOG_INFO(", (%8.2f)\n", device.currentAllocatedSize / 1024.0 / 1024.0);
#endif
ggml_backend_metal_log_allocated_size(device);

return ggml_backend_buffer_init(ggml_backend_metal_buffer_type(), ggml_backend_metal_buffer_i, ctx, size);
}
Expand Down