dhrystone: create nicer custom build, baremetal almost working

.gitmodules

@@ -14,6 +14,9 @@
 	path = submodules/crosstool-ng
     # url = https://github.com/crosstool-ng/crosstool-ng
 	url = https://github.com/cirosantilli/crosstool-ng
+[submodule "submodules/dhrystone"]
+	path = submodules/dhrystone
+	url = https://github.com/cirosantilli/dhrystone
 [submodule "submodules/gcc"]
 	path = submodules/gcc
     # url = git://gcc.gnu.org/git/gcc.git

README.adoc

@@ -3957,7 +3957,9 @@ Source: link:userland/c/getchar.c[]
 
 Let's see if user mode runs considerably faster than full system or not.
 
-First we build Dhrystone manually statically since dynamic linking is broken in gem5 as explained at: xref:gem5-syscall-emulation-mode[xrefstyle=full].
+First we build <<dhrystone>> manually statically since dynamic linking is broken in gem5 as explained at: xref:gem5-syscall-emulation-mode[xrefstyle=full].
+
+TODO: move this section to our new custom dhrystone setup: xref:dhrystone[xrefstyle=full].
 
 gem5 user mode:
 
@@ -10454,7 +10456,7 @@ gem5 however has tended towards intensive code generation in order to support al
 
 OK, this is why we used gem5 in the first place, performance measurements!
 
-Let's see how many cycles https://en.wikipedia.org/wiki/Dhrystone[Dhrystone], which Buildroot provides, takes for a few different input parameters.
+Let's see how many cycles <<dhrystone>>, which Buildroot provides, takes for a few different input parameters.
 
 We will do that for various input parameters on full system by taking a checkpoint after the boot finishes a fast atomic CPU boot, and then we will restore in a more detailed mode and run the benchmark:
 
@@ -10937,6 +10939,52 @@ Buildroot built-in libraries, mostly under Libraries > Other:
 
 There are not yet enabled, but it should be easy to so, see: xref:add-new-buildroot-packages[xrefstyle=full]
 
+===== Dhrystone
+
+https://en.wikipedia.org/wiki/Dhrystone
+
+Created in the 80's, it is not a representative measure of performance in modern computers anymore. It has mostly been replaced by https://en.wikipedia.org/wiki/SPECint[SPEC], which is... closed source! Unbelievable.
+
+<<buildroot>> has a `dhrystone` package, but because it is so interesting to us, we decided to also build it ourselves, which allows things like static and baremetal compilation more easily.
+
+Build and run on QEMU <<user-mode-simulation>>:
+
+....
+git submodule update --init submodules/dhrystone
+./build-dhrystone --mode userland
+./run --userland "$(./getvar userland_build_dir)/submodules/dhrystone/dhrystone"
+....
+
+Build and run on gem5 use mode:
+
+....
+./build-dhrystone --mode userland --static --force-rebuild
+./run --emulator gem5 --userland "$(./getvar userland_build_dir)/submodules/dhrystone/dhrystone"
+....
+
+TODO automate run more nicely.
+
+Build for <<baremetal>> execution and run it in baremetal QEMU:
+
+....
+./build-dhrystone --arch aarch64 --mode baremetal
+./run --arch aarch64 --baremetal "$(./getvar baremetal_build_dir)/submodules/dhrystone/dhrystone"
+....
+
+TODO: fix the build, just need to factor out all run arguments from link:build-baremetal[] into link:common.py[] and it should just work, no missing syscalls.
+
+If you really want the Buildroot package for some reason, build it with:
+
+....
+./build-buildroot --config 'BR2_PACKAGE_DHRYSTONE=y'
+....
+
+and run inside the guest from `PATH` with:
+
+....
+dhrystone
+....
+
 ===== BST vs heap vs hashmap
 
 The following benchmark setup works both:

baremetal/lib/syscalls.c

@@ -1,15 +1,21 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <sys/stat.h>
+#include <sys/times.h>
 
 #include <lkmc.h>
 
+#define UART_DR(baseaddr) (*(unsigned int *)(baseaddr))
+#define UART_FR(baseaddr) (*(((unsigned int *)(baseaddr))+6))
+
 enum {
     UART_FR_RXFE = 0x10,
 };
 
-#define UART_DR(baseaddr) (*(unsigned int *)(baseaddr))
-#define UART_FR(baseaddr) (*(((unsigned int *)(baseaddr))+6))
+extern char heap_low;
+extern char heap_top;
+
+char *heap_end = 0;
 
 void lkmc_baremetal_on_exit_callback(int status, void *arg) {
     (void)arg;
@@ -73,10 +79,15 @@ int _read(int file, char *ptr, int len) {
     return todo;
 }
 
-char *heap_end = 0;
+/* Dummy implementation that just increments an integer. */
+_CLOCK_T_ _times_r (struct _reent *r, struct tms *ptms) {
+    static long long unsigned t = 0;
+    (void)r;
+    (void)ptms;
+    return t++;
+}
+
 caddr_t _sbrk(int incr) {
-    extern char heap_low;
-    extern char heap_top;
     char *prev_heap_end;
     if (heap_end == 0) {
         heap_end = &heap_low;

build-baremetal

@@ -36,23 +36,6 @@ Build the baremetal examples with crosstool-NG.
             self.env['baremetal_build_lib_dir'],
             self.env['common_basename_noext'] + self.env['obj_ext']
         )
-        syscalls_basename_noext = 'syscalls'
-        syscalls_src = os.path.join(
-            self.env['baremetal_source_lib_dir'],
-            syscalls_basename_noext + self.env['c_ext']
-        )
-        syscalls_obj = os.path.join(
-            self.env['baremetal_build_lib_dir'],
-            syscalls_basename_noext + self.env['obj_ext']
-        )
-        syscalls_asm_src = os.path.join(
-            self.env['baremetal_source_lib_dir'],
-            syscalls_basename_noext + '_asm' + self.env['asm_ext']
-        )
-        syscalls_asm_obj = os.path.join(
-            self.env['baremetal_build_lib_dir'],
-            syscalls_basename_noext + '_asm' + self.env['obj_ext']
-        )
         cc_flags = [
             '-I', self.env['root_dir'], LF,
             '-O{}'.format(self.env['optimization_level']), LF,
@@ -101,8 +84,14 @@ Build the baremetal examples with crosstool-NG.
         for in_path, out_path in [
             (bootloader_src, extra_obj_baremetal_bootloader),
             (self.env['common_c'], extra_obj_lkmc_common),
-            (syscalls_src, syscalls_obj),
-            (syscalls_asm_src, syscalls_asm_obj),
+            (
+                self.env['baremetal_syscalls_src'],
+                self.env['baremetal_syscalls_obj']
+            ),
+            (
+                self.env['baremetal_syscalls_asm_src'],
+                self.env['baremetal_syscalls_asm_obj']
+            ),
         ]:
             self._build_one(
                 in_path=in_path,
@@ -133,7 +122,10 @@ Build the baremetal examples with crosstool-NG.
                                 self.env['baremetal_link_script'],
                                 self.env['common_h']
                             ],
-                            'extra_objs': [syscalls_obj, syscalls_asm_obj],
+                            'extra_objs': [
+                                self.env['baremetal_syscalls_obj'],
+                                self.env['baremetal_syscalls_asm_obj']
+                            ],
                             'extra_objs_baremetal_bootloader': [extra_obj_baremetal_bootloader],
                             'extra_objs_lkmc_common': [extra_obj_lkmc_common],
                             'in_path': in_path,

build-dhrystone

@@ -0,0 +1,62 @@
+#!/usr/bin/env python3
+
+import os
+import shutil
+
+import common
+import shlex
+from shell_helpers import LF
+
+class Main(common.BuildCliFunction):
+    def __init__(self):
+        super().__init__(
+            description='''\
+https://cirosantilli.com/linux-kernel-module-cheat#dhrystone
+'''
+        )
+        self._add_argument('--ccflags')
+        self._add_argument('--force-rebuild')
+        self._add_argument('--optimization-level')
+
+    def setup(self, env):
+        self.root_relpath = os.path.join('submodules', 'dhrystone')
+
+    def build(self):
+        build_dir = self.get_build_dir()
+        cflags = ['-O{}'.format(self.env['optimization_level'])]
+        extra_flags = []
+        if self.env['static']:
+            cflags.extend(['-static'])
+        if self.env['force_rebuild']:
+            extra_flags.extend(['-B', LF])
+        if self.env['mode'] == 'baremetal':
+            extra_objs = [
+                self.env['baremetal_syscalls_obj'],
+                self.env['baremetal_syscalls_asm_obj']
+            ]
+        else:
+            extra_objs = []
+        ret = self.sh.run_cmd(
+            [
+                'make', LF,
+                '-j', str(self.env['nproc']), LF,
+                '-C', os.path.join(self.env['submodules_dir'], 'dhrystone'), LF,
+                'CC={}'.format(self.env['gcc_path']), LF,
+                'CFLAGS={}'.format(' '.join(cflags)), LF,
+                'EXTRA_OBJS={}'.format(' '.join(extra_objs)), LF,
+                'OUT_DIR={}'.format(build_dir), LF,
+            ]
+            + extra_flags
+        )
+        if ret == 0 and self.env['mode'] == 'userland':
+            self.sh.copy_file_if_update(
+                os.path.join(build_dir, 'dhrystone'),
+                os.path.join(self.env['out_rootfs_overlay_lkmc_dir'], self.root_relpath),
+            )
+        return ret
+
+    def get_build_dir(self):
+        return os.path.join(self.env['build_dir'], self.root_relpath)
+
+if __name__ == '__main__':
+    Main().cli()

common.py

@@ -947,6 +947,23 @@ Incompatible archs are skipped.
             env['simulator_name'] = 'qemu'
         env['baremetal_build_dir'] = join(env['out_dir'], 'baremetal', env['arch'], env['simulator_name'], env['machine'])
         env['baremetal_build_lib_dir'] = join(env['baremetal_build_dir'], env['baremetal_lib_basename'])
+        env['baremetal_syscalls_basename_noext'] = 'syscalls'
+        env['baremetal_syscalls_src'] = os.path.join(
+            env['baremetal_source_lib_dir'],
+            env['baremetal_syscalls_basename_noext'] + self.env['c_ext']
+        )
+        env['baremetal_syscalls_obj'] = os.path.join(
+            self.env['baremetal_build_lib_dir'],
+            env['baremetal_syscalls_basename_noext'] + self.env['obj_ext']
+        )
+        env['baremetal_syscalls_asm_src'] = os.path.join(
+            self.env['baremetal_source_lib_dir'],
+            env['baremetal_syscalls_basename_noext'] + '_asm' + self.env['asm_ext']
+        )
+        env['baremetal_syscalls_asm_obj'] = os.path.join(
+            self.env['baremetal_build_lib_dir'],
+            env['baremetal_syscalls_basename_noext'] + '_asm' + self.env['obj_ext']
+        )
 
         # Userland / baremetal common source.
         env['common_basename_noext'] = env['repo_short_id']
@@ -958,6 +975,10 @@ Incompatible archs are skipped.
             env['root_dir'],
             env['common_basename_noext'] + env['header_ext']
         )
+        if env['mode'] == 'baremetal':
+            env['build_dir'] = env['baremetal_build_dir']
+        elif env['mode'] == 'userland':
+            env['build_dir'] = env['userland_build_dir']
 
         # Docker
         env['docker_build_dir'] = join(env['out_dir'], 'docker', env['arch'])
@@ -1022,8 +1043,10 @@ lunch aosp_{}-eng
 '''.format(self.env['android_arch'])
 
         # Toolchain.
+        if env['baremetal'] and not env['_args_given']['mode']:
+            env['mode'] = 'baremetal'
         if not env['_args_given']['gcc_which']:
-            if env['baremetal']:
+            if env['mode'] == 'baremetal':
                 env['gcc_which'] = 'crosstool-ng'
         if env['gcc_which'] == 'buildroot':
             env['toolchain_prefix'] = os.path.join(

shell_helpers.py

@@ -162,6 +162,15 @@ class ShellHelpers:
             ending = last_newline + ';'
         return newline_separator.join(out) + ending
 
+    def copy_file_if_update(self, src, dest):
+        if os.path.isdir(dest):
+            dest = os.path.join(dest, os.path.basename(src))
+        if (
+            not os.path.exists(dest) or \
+            os.path.getmtime(src) > os.path.getmtime(dest)
+        ):
+            self.cp(src, dest)
+
     def copy_dir_if_update_non_recursive(
         self,
         srcdir,
@@ -178,12 +187,9 @@ class ShellHelpers:
             src = os.path.join(srcdir, basename)
             if os.path.isfile(src) or os.path.islink(src):
                 noext, ext = os.path.splitext(basename)
-                dest = os.path.join(destdir, basename)
-                if (
-                    (filter_ext is None or ext == filter_ext) and
-                    (not os.path.exists(dest) or os.path.getmtime(src) > os.path.getmtime(dest))
-                ):
-                    self.cp(src, dest)
+                if (filter_ext is None or ext == filter_ext):
+                    dest = os.path.join(destdir, basename)
+                    self.copy_file_if_update(src, dest)
 
     def copy_dir_if_update(
         self,