Matthew Poremba has uploaded this change for review. (
https://gem5-review.googlesource.com/c/public/gem5/+/58472 )
Change subject: gpu-compute: Fatal on dynamic scratch allocation in GPUFS
......................................................................
gpu-compute: Fatal on dynamic scratch allocation in GPUFS
This is known not working in GPUFS. As a result, the simulation will
never end. Rather than simulate forever, add a fatal for now to exit
simulation until support for this functionality is added.
Change-Id: I8e45996a7eb781575e8643baea05daf87bc5f1c3
---
M src/gpu-compute/gpu_command_processor.hh
1 file changed, 18 insertions(+), 0 deletions(-)
diff --git a/src/gpu-compute/gpu_command_processor.hh
b/src/gpu-compute/gpu_command_processor.hh
index ec163ed..ba8b007 100644
--- a/src/gpu-compute/gpu_command_processor.hh
+++ b/src/gpu-compute/gpu_command_processor.hh
@@ -57,6 +57,7 @@
#include "gpu-compute/gpu_compute_driver.hh"
#include "gpu-compute/hsa_queue_entry.hh"
#include "params/GPUCommandProcessor.hh"
+#include "sim/full_system.hh"
namespace gem5
{
@@ -219,6 +220,10 @@
task->amdQueue.compute_tmpring_size_wavesize * 1024,
task->privMemPerItem());
+ // Currently this is not supported in GPU full system
+ fatal_if(FullSystem,
+ "Runtime dynamic scratch allocation not supported");
+
updateHsaSignal(task->amdQueue.queue_inactive_signal.handle, 1,
[ = ] (const uint64_t &dma_buffer)
{ WaitScratchDmaEvent(task, dma_buffer);
});
--
To view, visit https://gem5-review.googlesource.com/c/public/gem5/+/58472
To unsubscribe, or for help writing mail filters, visit
https://gem5-review.googlesource.com/settings
Gerrit-Project: public/gem5
Gerrit-Branch: develop
Gerrit-Change-Id: I8e45996a7eb781575e8643baea05daf87bc5f1c3
Gerrit-Change-Number: 58472
Gerrit-PatchSet: 1
Gerrit-Owner: Matthew Poremba <matthew.pore...@amd.com>
Gerrit-MessageType: newchange
_______________________________________________
gem5-dev mailing list -- gem5-dev@gem5.org
To unsubscribe send an email to gem5-dev-le...@gem5.org
%(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
