AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Support more than one bundles in a single .kernel section

Browse Source 
When compiling with Early Finalization enabled in HCC,
the resulting .kernel section of the host object now may
contain more than one device code bundles.  This is to
teach the HIP runtime to correctly extract all the
bundles from the .kernel section.


[ROCm/clr commit: f39873c584]
This commit is contained in:
Siu Chi Chan
2018-10-18 16:53:03 -04:00
parent 1e6207db23
commit 5d1a265efc
2 changed files with 15 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/clr/hipamd/src/program_state.cpp
+10 -4
View File
@@ -209,10 +209,16 @@ const unordered_map<hsa_isa_t, vector<vector<char>>>& code_object_blobs(bool reb
nullptr);
for (auto&& blob : blobs) {
Bundled_code_header tmp{blob};
if (valid(tmp)) {
for (auto&& bundle : bundles(tmp)) {
r[triple_to_hsa_isa(bundle.triple)].push_back(bundle.blob);
for (auto sub_blob = blob.begin(); sub_blob != blob.end(); ) {
Bundled_code_header tmp(sub_blob, blob.end());
if (valid(tmp)) {
for (auto&& bundle : bundles(tmp)) {
r[triple_to_hsa_isa(bundle.triple)].push_back(bundle.blob);
}
sub_blob+=tmp.bundled_code_size;
}
else {
break;
}
}
}