Code object compression via bundling

[bstefanuk]

Summary:

This PR adds a compression layer to all final code objects, thereby generating smaller libraries at the expense of build time. Includes minor refactoring.

Outcomes:

• A new clang-offload-bundler invocation is added after assembly object linking.
• getAssemblyCodeObjectFiles has been renamed to buildAssemblyCodeObjectFiles to match the name of source kernel functions.

Build	Time	Build Size (build/library/ directory)
Feature, gfx90a	8m38.902s	269M
Develop, gfx90a	8m3.921s	812M
Feature, gfx90a		2.4G
Develop, gfx90a		11G

Testing and Environment:

Docker: Ubuntu 24.04, ROCm 6.4 RC stack, AMD clang version 18.0.0, AMD clang-offload-bundler version 18.0.0

Tested with hipBLASLt bench and test clients

[TorreZuk]

Basics look good, assume all rocblas tests passed.

[TorreZuk]

Can you consider doing what I added for windows (response file) instead of processing in batches? Is there a limit of 10000 with the linker?

[bstefanuk]

In TensileLite, this batching strategy is implemented, but the source PR doesn't describe the motivation beyond your interpretation: some error was faced in the past because too many arguments were issued to the compiler.

I spent some time trying to find documentation on this 10000-input limit but couldn't find anything meaningful, so I left the implementation intact. @KKyang I believe that you implemented this feature, could you provide more context?

[KKyang]

#415 this?

[bstefanuk]

That's correct.

[LunNova]

Might be worth zstd compressing the msgpack files too, they're pretty compressible. Here's an untested attempt at decompress support in case it's helpful:

diff --git a/Tensile/Source/lib/source/msgpack/MessagePack.cpp b/Tensile/Source/lib/source/msgpack/MessagePack.cpp
index de97929c..dbc397e0 100644
--- a/Tensile/Source/lib/source/msgpack/MessagePack.cpp
+++ b/Tensile/Source/lib/source/msgpack/MessagePack.cpp
@@ -28,6 +28,8 @@
 
 #include <Tensile/msgpack/Loading.hpp>
 
+#include <zstd.h>
+
 #include <fstream>
 
 namespace Tensile
@@ -86,6 +88,34 @@ namespace Tensile
                 return nullptr;
             }
 
+            // Check if the file is zstd compressed
+            char magic[4];
+            in.read(magic, 4);
+            bool isCompressed = (in.gcount() == 4 && magic[0] == '\x28' && magic[1] == '\xB5' && magic[2] == '\x2F' && magic[3] == '\xFD');
+            // Reset file pointer to the beginning
+            in.seekg(0, std::ios::beg);
+
+            if (isCompressed) {
+                // Decompress zstd file
+                std::vector<char> compressedData((std::istreambuf_iterator<char>(in)), std::istreambuf_iterator<char>());
+
+                size_t decompressedSize = ZSTD_getFrameContentSize(compressedData.data(), compressedData.size());
+                if (decompressedSize == ZSTD_CONTENTSIZE_ERROR || decompressedSize == ZSTD_CONTENTSIZE_UNKNOWN) {
+                    if(Debug::Instance().printDataInit())
+                        std::cout << "Error: Unable to determine decompressed size for " << filename << std::endl;
+                    return nullptr;
+                }
+
+                std::vector<char> decompressedData(decompressedSize);
+                size_t dSize = ZSTD_decompress(decompressedData.data(), decompressedSize, compressedData.data(), compressedData.size());
+                if (ZSTD_isError(dSize)) {
+                    if(Debug::Instance().printDataInit())
+                        std::cout << "Error: ZSTD decompression failed for " << filename << std::endl;
+                    return nullptr;
+                }
+
+                msgpack::unpack(result, decompressedData.data(), dSize);
+            } else {
             msgpack::unpacker unp;
             bool              finished_parsing;
             constexpr size_t  buffer_size = 1 << 19;
@@ -109,6 +139,7 @@ namespace Tensile
 
                 return nullptr;
             }
+            }
         }
         catch(std::runtime_error const& exc)
         {

[bstefanuk]

@LunNova Thanks for the code snippet. I've had this idea as well and have plans to implement it. However, for the scope of this PR we'll keep it to code object files and add the .dat file compression in another PR.