[Lldb-commits] [lldb] [lldb][RISCV] Handle subsets of CSRs in RV32 core dump images (PR #142932)
Jonas Devlieghere via lldb-commits
lldb-commits at lists.llvm.org
Fri Mar 6 09:10:06 PST 2026
================
@@ -7,75 +7,372 @@
//===----------------------------------------------------------------------===//
#include "RegisterContextPOSIXCore_riscv32.h"
+
#include "lldb/Utility/DataBufferHeap.h"
+#define GPR_OFFSET(idx) ((idx) * sizeof(uint32_t))
+#define FPR_OFFSET(idx) ((idx) * sizeof(uint32_t))
+#define CSR_OFFSET(idx) ((idx) * sizeof(uint32_t))
+
+#define DECLARE_REGISTER_INFOS_RISCV32_STRUCT
+#include "Plugins/Process/Utility/RegisterInfos_riscv32.h"
+#undef DECLARE_REGISTER_INFOS_RISCV32_STRUCT
+
using namespace lldb_private;
std::unique_ptr<RegisterContextCorePOSIX_riscv32>
RegisterContextCorePOSIX_riscv32::Create(Thread &thread, const ArchSpec &arch,
const DataExtractor &gpregset,
llvm::ArrayRef<CoreNote> notes) {
- Flags opt_regsets = RegisterInfoPOSIX_riscv32::eRegsetMaskDefault;
-
return std::unique_ptr<RegisterContextCorePOSIX_riscv32>(
new RegisterContextCorePOSIX_riscv32(
- thread,
- std::make_unique<RegisterInfoPOSIX_riscv32>(arch, opt_regsets),
+ thread, std::make_unique<RegisterInfoPOSIXDynamic_riscv32>(arch),
gpregset, notes));
}
RegisterContextCorePOSIX_riscv32::RegisterContextCorePOSIX_riscv32(
- Thread &thread, std::unique_ptr<RegisterInfoPOSIX_riscv32> register_info,
+ Thread &thread,
+ std::unique_ptr<RegisterInfoPOSIXDynamic_riscv32> register_info,
const DataExtractor &gpregset, llvm::ArrayRef<CoreNote> notes)
- : RegisterContextPOSIX_riscv32(thread, std::move(register_info)) {
+ : RegisterContext(thread, 0), m_reg_infos_up(std::move(register_info)) {
+ // Compute the maximum register counts for GPR, FPR, and CSR.
+ uint32_t k_num_gpr_registers = (sizeof(g_register_infos_riscv32_gpr) /
----------------
JDevlieghere wrote:
Assuming `g_register_infos_riscv32_gpr` is `constexpr`, these should be as well, especially given their `k_` prefix which suggests they're constants but the code allows them to be mutated.
https://github.com/llvm/llvm-project/pull/142932
More information about the lldb-commits
mailing list