[build] Allow for 32-bit and 64-bit versions of util/zbin

Parsing ELF data is simpler if we don't have to build a single binary
to handle both 32-bit and 64-bit ELF formats.

Allow for separate 32-bit and 64-bit binaries built from util/zbin.c
(as is already done for util/elf2efi.c).

Signed-off-by: Michael Brown <mcb30@ipxe.org>

src/arch/arm32/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN32)
+
 # ARM32-specific directories containing source files
 #
 SRCDIRS		+= arch/arm32/core

src/arch/arm64/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN64)
+
 # ARM64-specific directories containing source files
 #
 SRCDIRS		+= arch/arm64/core

src/arch/i386/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN32)
+
 # Force i386-only instructions
 #
 CFLAGS		+= -march=i386

src/arch/loong64/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN64)
+
 # Assembler section type character
 #
 ASM_TCHAR	:= @

src/arch/riscv32/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN32)
+
 # RISCV32-specific directories containing source files
 #
 SRCDIRS		+= arch/riscv32/core

src/arch/riscv64/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN64)
+
 # RISCV64-specific directories containing source files
 #
 SRCDIRS		+= arch/riscv64/core

src/arch/x86_64/Makefile

@@ -1,3 +1,7 @@
+# Specify compressor
+#
+ZBIN		= $(ZBIN64)
+
 # Code size reduction.
 #
 CFLAGS		+= -fstrength-reduce -fomit-frame-pointer