derek/gem5
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

mem-ruby: Fix deadlock in VIPERCoalescer

Browse Source 
Certain instructions (some atomics and buffer_wbinvl1_vol) deadlock
in the coalescer, where sendTimingReq fails, fails a retry, and then
never retries again.

This fix sets m_cache_inv_pkt to null before calling
completeHitCallback(), as that allows the failed packets to be retried
again.

Change-Id: I4a51c741360f385f8b4c3f2a31a9410f18e095d9
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/37477
Reviewed-by: Matt Sinclair <mattdsinclair@gmail.com>
Reviewed-by: Matthew Poremba <matthew.poremba@amd.com>
Maintainer: Matt Sinclair <mattdsinclair@gmail.com>
Tested-by: kokoro <noreply+kokoro@google.com>
This commit is contained in:
Kyle Roarty
2020-11-10 00:36:05 -06:00
parent 5f6ebe752e
commit a107f62b41
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/mem/ruby/system/VIPERCoalescer.cc
View File

@@ -269,8 +269,8 @@ VIPERCoalescer::invTCPCallback(Addr addr)
if (m_num_pending_invs == 0) {
std::vector<PacketPtr> pkt_list { m_cache_inv_pkt };
completeHitCallback(pkt_list);
m_cache_inv_pkt = nullptr;
completeHitCallback(pkt_list);
}
}