RISC-V: always report presence of extensions formerly part of the bas…

…e ISA

Of these four extensions, two were part of the base ISA when the port was
written and are required by the kernel. The other two are implied when
`i` is in riscv,isa on DT systems.
There's not much that userspace can do with this extra information, but
there is no harm in reporting an ISA string that closer resembles the
current versions of the specifications either.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230607-nest-collision-5796b6be8be6@spud
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

arch/riscv/include/asm/hwcap.h

@@ -46,6 +46,10 @@
 #define RISCV_ISA_EXT_ZICBOZ		34
 #define RISCV_ISA_EXT_SMAIA		35
 #define RISCV_ISA_EXT_SSAIA		36
+#define RISCV_ISA_EXT_ZICNTR		37
+#define RISCV_ISA_EXT_ZICSR		38
+#define RISCV_ISA_EXT_ZIFENCEI		39
+#define RISCV_ISA_EXT_ZIHPM		40
 
 #define RISCV_ISA_EXT_MAX		64
 #define RISCV_ISA_EXT_NAME_LEN_MAX	32

arch/riscv/kernel/cpu.c

@@ -208,7 +208,11 @@ arch_initcall(riscv_cpuinfo_init);
 static struct riscv_isa_ext_data isa_ext_arr[] = {
 	__RISCV_ISA_EXT_DATA(zicbom, RISCV_ISA_EXT_ZICBOM),
 	__RISCV_ISA_EXT_DATA(zicboz, RISCV_ISA_EXT_ZICBOZ),
+	__RISCV_ISA_EXT_DATA(zicntr, RISCV_ISA_EXT_ZICNTR),
+	__RISCV_ISA_EXT_DATA(zicsr, RISCV_ISA_EXT_ZICSR),
+	__RISCV_ISA_EXT_DATA(zifencei, RISCV_ISA_EXT_ZIFENCEI),
 	__RISCV_ISA_EXT_DATA(zihintpause, RISCV_ISA_EXT_ZIHINTPAUSE),
+	__RISCV_ISA_EXT_DATA(zihpm, RISCV_ISA_EXT_ZIHPM),
 	__RISCV_ISA_EXT_DATA(zbb, RISCV_ISA_EXT_ZBB),
 	__RISCV_ISA_EXT_DATA(smaia, RISCV_ISA_EXT_SMAIA),
 	__RISCV_ISA_EXT_DATA(ssaia, RISCV_ISA_EXT_SSAIA),

arch/riscv/kernel/cpufeature.c

@@ -311,6 +311,23 @@ void __init riscv_fill_hwcap(void)
 #undef SET_ISA_EXT_MAP
 		}
 
+		/*
+		 * Linux requires the following extensions, so we may as well
+		 * always set them.
+		 */
+		set_bit(RISCV_ISA_EXT_ZICSR, this_isa);
+		set_bit(RISCV_ISA_EXT_ZIFENCEI, this_isa);
+
+		/*
+		 * These ones were as they were part of the base ISA when the
+		 * port & dt-bindings were upstreamed, and so can be set
+		 * unconditionally where `i` is in riscv,isa on DT systems.
+		 */
+		if (acpi_disabled) {
+			set_bit(RISCV_ISA_EXT_ZICNTR, this_isa);
+			set_bit(RISCV_ISA_EXT_ZIHPM, this_isa);
+		}
+
 		/*
 		 * All "okay" hart should have same isa. Set HWCAP based on
 		 * common capabilities of every "okay" hart, in case they don't