*/
-#ifdef __SHARC__
-#pragma nosharc
-#endif
-
#include <ros/common.h>
#include <smp.h>
#include <arch/init.h>
#include <mm.h>
#include <elf.h>
-#include <frontend.h>
#include <kmalloc.h>
#include <assert.h>
#include <process.h>
#include <schedule.h>
#include <syscall.h>
-#include <testing.h>
+#include <ktest.h>
#include <kfs.h>
#include <stdio.h>
-#include <timing.h>
-#include <resource.h>
+#include <time.h>
#include <monitor.h>
-#include <colored_caches.h>
#include <string.h>
#include <pmap.h>
-#include <ros/timer.h>
+#include <arch/console.h>
+#include <time.h>
#include <ros/arch/membar.h>
/*
*/
void manager(void)
{
- #ifndef DEVELOPER_NAME
- #define DEVELOPER_NAME brho
- #endif
-
// LoL
#define PASTE(s1,s2) s1 ## s2
#define MANAGER_FUNC(dev) PASTE(manager_,dev)
+ #if !defined(DEVELOPER_NAME) && \
+ (defined(CONFIG_KERNEL_TESTING) || \
+ defined(CONFIG_USERSPACE_TESTING))
+ #define DEVELOPER_NAME jenkins
+ #endif
+
+ #ifndef DEVELOPER_NAME
+ #define DEVELOPER_NAME brho
+ #endif
+
void MANAGER_FUNC(DEVELOPER_NAME)(void);
MANAGER_FUNC(DEVELOPER_NAME)();
}
-/* Helper macro for quickly running something out of KFS. Pass it a string and
- * a proc pointer. */
-#define quick_proc_run(x, p) \
- (p) = kfs_proc_create(kfs_lookup_path((x))); \
+char *p_argv[] = {0, 0, 0};
+/* Helper macro for quickly running a process. Pass it a string, *file, and a
+ * *proc. */
+#define quick_proc_run(x, p, f) \
+ (f) = do_file_open((x), O_READ, 0); \
+ assert((f)); \
+ p_argv[0] = file_name((f)); \
+ (p) = proc_create((f), p_argv, NULL); \
+ kref_put(&(f)->f_kref); \
spin_lock(&(p)->proc_lock); \
__proc_set_state((p), PROC_RUNNABLE_S); \
spin_unlock(&(p)->proc_lock); \
- proc_run((p)); \
- proc_decref((p), 1);
-
-#define quick_proc_create(x, p) \
- (p) = kfs_proc_create(kfs_lookup_path((x))); \
+ proc_run_s((p)); \
+ proc_decref((p));
+
+#define quick_proc_create(x, p, f) \
+ (f) = do_file_open((x), O_READ, 0); \
+ assert((f)); \
+ p_argv[0] = file_name((f)); \
+ (p) = proc_create((f), p_argv, NULL); \
+ kref_put(&(f)->f_kref); \
spin_lock(&(p)->proc_lock); \
__proc_set_state((p), PROC_RUNNABLE_S); \
spin_unlock(&(p)->proc_lock);
-#define quick_proc_color_run(x, p, c) \
- (p) = kfs_proc_create(kfs_lookup_path((x))); \
+#define quick_proc_color_run(x, p, c, f) \
+ (f) = do_file_open((x), O_READ, 0); \
+ assert((f)); \
+ p_argv[0] = file_name((f)); \
+ (p) = proc_create((f), p_argv, NULL); \
+ kref_put(&(f)->f_kref); \
spin_lock(&(p)->proc_lock); \
__proc_set_state((p), PROC_RUNNABLE_S); \
spin_unlock(&(p)->proc_lock); \
p->cache_colors_map = cache_colors_map_alloc(); \
for (int i = 0; i < (c); i++) \
cache_color_alloc(llc_cache, p->cache_colors_map); \
- proc_run((p)); \
- proc_decref((p), 1);
-
-#define quick_proc_color_create(x, p, c) \
- (p) = kfs_proc_create(kfs_lookup_path((x))); \
+ proc_run_s((p)); \
+ proc_decref((p));
+
+#define quick_proc_color_create(x, p, c, f) \
+ (f) = do_file_open((x), O_READ, 0); \
+ assert((f)); \
+ p_argv[0] = file_name((f)); \
+ (p) = proc_create((f), p_argv, NULL); \
+ kref_put(&(f)->f_kref); \
spin_lock(&(p)->proc_lock); \
__proc_set_state((p), PROC_RUNNABLE_S); \
spin_unlock(&(p)->proc_lock); \
void manager_brho(void)
{
- static uint8_t RACY progress = 0;
- static struct proc *p;
-
+ static bool first = TRUE;
+ struct per_cpu_info *pcpui = &per_cpu_info[core_id()];
+
+ if (first) {
+ printk("*** IRQs must be enabled for input emergency codes ***\n");
+ #ifdef CONFIG_X86
+ printk("*** Hit ctrl-g to enter the monitor. ***\n");
+ printk("*** Hit ctrl-q to force-enter the monitor. ***\n");
+ printk("*** Hit ctrl-b for a backtrace of core 0 ***\n");
+ #else
+ printk("*** Hit ctrl-g to enter the monitor. ***\n");
+ #warning "***** ctrl-g untested on riscv, check k/a/r/trap.c *****"
+ #endif
+ first = FALSE;
+ }
+ /* just idle, and deal with things via interrupts. or via face. */
+ smp_idle();
+ /* whatever we do in the manager, keep in mind that we need to not do
+ * anything too soon (like make processes), since we'll drop in here during
+ * boot if the boot sequence required any I/O (like EXT2), and we need to
+ * PRKM() */
+ assert(0);
+
+#if 0 /* ancient tests below: (keeping around til we ditch the manager) */
// for testing taking cores, check in case 1 for usage
- uint32_t corelist[MAX_NUM_CPUS];
+ uint32_t corelist[MAX_NUM_CORES];
uint32_t num = 3;
+ struct file *temp_f;
+ static struct proc *p;
+ static uint8_t RACY progress = 0; /* this will wrap around. */
switch (progress++) {
case 0:
+ printk("Top of the manager to ya!\n");
/* 124 is half of the available boxboro colors (with the kernel
* getting 8) */
- quick_proc_color_run("msr_dumb_while", p, 124);
- //quick_proc_run("msr_dumb_while", p);
+ //quick_proc_color_run("msr_dumb_while", p, 124, temp_f);
+ quick_proc_run("/bin/hello", p, temp_f);
#if 0
// this is how you can transition to a parallel process manually
// make sure you don't proc run first
envs[0] = kfs_proc_create(kfs_lookup_path("roslib_hello"));
__proc_set_state(envs[0], PROC_RUNNABLE_S);
proc_run(envs[0]);
+ warn("DEPRECATED");
break;
#endif
case 2:
printd("Manager Progress: %d\n", progress);
// delay if you want to test rescheduling an MCP that yielded
//udelay(15000000);
- schedule();
+ run_scheduler();
}
panic("If you see me, then you probably screwed up");
monitor(0);
}
*/
return;
+#endif
}
-void manager_klueska()
-{
- static struct proc *envs[256];
- static volatile uint8_t progress = 0;
-
- if (progress == 0) {
- progress++;
- envs[0] = kfs_proc_create(kfs_lookup_path("fillmeup"));
- __proc_set_state(envs[0], PROC_RUNNABLE_S);
- proc_run(envs[0]);
- }
- schedule();
-
- panic("DON'T PANIC");
-}
-
-struct elf_info
+void manager_jenkins()
{
- long entry;
- long phdr;
- int phnum;
- int dynamic;
- char interp[256];
-};
-
-void manager_waterman()
-{
- static int init = 0;
- if(!init)
- {
- init = 1;
- struct proc* p = proc_create(NULL,0);
-
- char* argv[] = {"/bin/sh","-l",0};
- char* envp[] = {"LD_LIBRARY_PATH=/lib",0};
- procinfo_pack_args(p->procinfo,argv,envp);
-
- struct file* f = file_open("/bin/busybox",0,0);
- assert(f != NULL);
- assert(load_elf(p,f) == 0);
- file_decref(f);
+ #ifdef CONFIG_KERNEL_TESTING
+ printk("<-- BEGIN_KERNEL_TESTS -->\n");
+ run_registered_ktest_suites();
+ printk("<-- END_KERNEL_TESTS -->\n");
+ #endif
- __proc_set_state(p, PROC_RUNNABLE_S);
- proc_run(p);
+ // Run userspace tests (from config specified path).
+ #ifdef CONFIG_USERSPACE_TESTING
+ if (strlen(CONFIG_USERSPACE_TESTING_SCRIPT) != 0) {
+ char exec[] = "/bin/ash";
+ char *p_argv[] = {exec, CONFIG_USERSPACE_TESTING_SCRIPT, 0};
+
+ struct file *program = do_file_open(exec, O_READ, 0);
+ struct proc *p = proc_create(program, p_argv, NULL);
+ proc_wakeup(p);
+ proc_decref(p); /* let go of the reference created in proc_create() */
+ kref_put(&program->f_kref);
+ run_scheduler();
+ // Need a way to wait for p to finish
+ } else {
+ printk("No user-space launcher file specified.\n");
}
- schedule();
+ #endif
+ smp_idle();
+ assert(0);
}
-void manager_pearce()
+void manager_klueska()
{
static struct proc *envs[256];
static volatile uint8_t progress = 0;
if (progress == 0) {
progress++;
- envs[0] = kfs_proc_create(kfs_lookup_path("parlib_httpserver_integrated"));
- //envs[0] = kfs_proc_create(kfs_lookup_path("parlib_lock_test"));
+ panic("what do you want to do?");
+ //envs[0] = kfs_proc_create(kfs_lookup_path("fillmeup"));
__proc_set_state(envs[0], PROC_RUNNABLE_S);
- proc_run(envs[0]);
+ proc_run_s(envs[0]);
+ warn("DEPRECATED");
}
- schedule();
+ run_scheduler();
panic("DON'T PANIC");
-
-}
-
-#ifdef __CONFIG_OSDI__
-/* Manager for Micro benchmarks, OSDI, etc */
-struct proc *mgr_p1 = 0;
-struct proc *mgr_p2 = 0;
-static void exper_1_part2(struct proc **pp);
-static void exper_2_part2(struct proc **pp);
-static void exper_3_part2(struct proc **pp);
-static void exper_4_part2(struct proc **pp);
-static void exper_5_part2(struct proc **pp);
-static void exper_6_part2(struct proc **pp);
-static void exper_7_part2(struct proc **pp);
-static void exper_8_part2(struct proc **pp);
-static void exper_9_part2(struct proc **pp);
-
-void manager_tests(void)
-{
- static uint8_t RACY progress = 0;
-
- printk("Test Progress: %d\n", progress);
- /* 10 runs of every experiment. Finishing/Part2 is harmless on a null
- * pointer. We need to clean up/ finish/ part2 after each quick_proc_run,
- * since we leave the monitor and only enter on another run (with
- * progress++). That's why we run a part2 in the first case: of the next
- * experiment. */
- switch (progress++) {
- /* Experiment 1: get max vcores */
- case 0:
- printk("************* Starting experiment 1 ************** \n");
- case 1:
- case 2:
- case 3:
- case 4:
- case 5:
- case 6:
- case 7:
- case 8:
- case 9:
- exper_1_part2(&mgr_p1);
- quick_proc_run("msr_get_cores", mgr_p1);
- break;
- /* Experiment 2: get a single vcore */
- case 10:
- exper_1_part2(&mgr_p1);
- printk("************* Starting experiment 2 ************** \n");
- case 11:
- case 12:
- case 13:
- case 14:
- case 15:
- case 16:
- case 17:
- case 18:
- case 19:
- exper_2_part2(&mgr_p1);
- quick_proc_run("msr_get_singlecore", mgr_p1);
- break;
- /* Experiment 3: kill a _M */
- case 20: /* leftover from exp 2 */
- exper_2_part2(&mgr_p1);
- printk("************* Starting experiment 3 ************** \n");
- case 21:
- case 22:
- case 23:
- case 24:
- case 25:
- case 26:
- case 27:
- case 28:
- case 29:
- exper_3_part2(&mgr_p1);
- quick_proc_run("msr_dumb_while", mgr_p1);
- break;
- /* Experiment 4: _S create and death*/
- case 30: /* leftover from exp 3 */
- exper_3_part2(&mgr_p1);
- printk("************* Starting experiment 4 ************** \n");
- case 31:
- case 32:
- case 33:
- case 34:
- case 35:
- case 36:
- case 37:
- case 38:
- case 39:
- exper_4_part2(&mgr_p1);
- printk("[T]:004:S:%llu\n", read_tsc());
- quick_proc_run("tsc_spitter", mgr_p1);
- break;
- /* Experiment 5: raw preempt, entire process*/
- case 40:
- exper_4_part2(&mgr_p1);
- printk("************* Starting experiment 5 ************** \n");
- case 41:
- case 42:
- case 43:
- case 44:
- case 45:
- case 46:
- case 47:
- case 48:
- case 49:
- exper_5_part2(&mgr_p1);
- quick_proc_run("msr_nice_while", mgr_p1);
- break;
- /* Experiment 6: preempt-warn, entire process */
- case 50:
- exper_5_part2(&mgr_p1);
- printk("************* Starting experiment 6 ************** \n");
- case 51:
- case 52:
- case 53:
- case 54:
- case 55:
- case 56:
- case 57:
- case 58:
- case 59:
- exper_6_part2(&mgr_p1);
- quick_proc_run("msr_nice_while", mgr_p1);
- break;
- /* Experiment 7: preempt-raw, single core */
- case 60:
- exper_6_part2(&mgr_p1);
- printk("************* Starting experiment 7 ************** \n");
- case 61:
- case 62:
- case 63:
- case 64:
- case 65:
- case 66:
- case 67:
- case 68:
- case 69:
- exper_7_part2(&mgr_p1);
- quick_proc_run("msr_nice_while", mgr_p1);
- break;
- /* Experiment 8: preempt-warn, single core */
- case 70:
- exper_7_part2(&mgr_p1);
- printk("************* Starting experiment 8 ************** \n");
- case 71:
- case 72:
- case 73:
- case 74:
- case 75:
- case 76:
- case 77:
- case 78:
- case 79:
- exper_8_part2(&mgr_p1);
- quick_proc_run("msr_nice_while", mgr_p1);
- break;
- /* Experiment 9: single notification time */
- case 80:
- exper_8_part2(&mgr_p1);
- printk("************* Starting experiment 9 ************** \n");
- case 81:
- case 82:
- case 83:
- case 84:
- case 85:
- case 86:
- case 87:
- case 88:
- case 89:
- exper_9_part2(&mgr_p1);
- quick_proc_run("msr_dumb_while", mgr_p1);
- break;
- /* Experiment 10: cycling vcore */
- case 90:
- exper_9_part2(&mgr_p1);
- printk("************* Starting experiment 10 ************* \n");
- quick_proc_run("msr_dumb_while", mgr_p1);
- break;
- case 91:
- quick_proc_run("msr_cycling_vcores", mgr_p2);
- break;
- case 92:
- printk("Will go on forever. Udelaying for two minutes.\n");
- udelay(120000000);
- proc_incref(mgr_p1, 1);
- proc_destroy(mgr_p1);
- proc_decref(mgr_p1, 1);
- proc_incref(mgr_p2, 1);
- proc_destroy(mgr_p2);
- proc_decref(mgr_p2, 1);
- printk("Done with the tests!");
- monitor(0);
- break;
- default:
- printd("Manager Progress: %d\n", progress);
- schedule();
- }
- monitor(0);
- return;
-}
-
-/* OSDI experiment "bottom halves" */
-/* Experiment 1: get max vcores */
-static void exper_1_part2(struct proc **pp)
-{
- while (*pp) /* make sure the previous run is over */
- cpu_relax();
-}
-
-/* Experiment 2: get a single vcore */
-static void exper_2_part2(struct proc **pp)
-{
- while (*pp) /* make sure the previous run is over */
- cpu_relax();
-}
-
-/* Experiment 3: kill a _M */
-static void exper_3_part2(struct proc **pp)
-{
- uint64_t begin = 0, diff = 0;
-
- if (*pp) { /* need to kill, etc */
- proc_incref(*pp, 1);
- begin = start_timing();
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- wmb();
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- diff = stop_timing(begin);
- printk("Took %llu usec (%llu nsec) to kill.\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- printk("[T]:003:%llu:%llu\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- }
-}
-
-/* Experiment 4: _S create and death*/
-static void exper_4_part2(struct proc **pp)
-{
- while (*pp) /* make sure the previous run is over */
- cpu_relax();
-}
-
-/* Experiment 5: raw preempt, entire process*/
-static void exper_5_part2(struct proc **pp)
-{
- uint64_t begin = 0, diff = 0;
- uint32_t end_refcnt = 0;
- bool self_ipi_pending = FALSE;
-
- if (*pp) {
- proc_incref(*pp, 1);
- spin_lock(&(*pp)->proc_lock);
- end_refcnt = (*pp)->env_refcnt - (*pp)->procinfo->num_vcores;
- begin = start_timing();
- self_ipi_pending = __proc_preempt_all(*pp);
- spin_unlock(&(*pp)->proc_lock);
- __proc_kmsg_pending(*pp, self_ipi_pending);
- spin_on((*pp)->env_refcnt != end_refcnt);
- diff = stop_timing(begin);
- printk("Took %llu usec (%llu nsec) to raw preempt all.\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- printk("[T]:005:%llu:%llu\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- }
-}
-
-/* Experiment 6: preempt-warn, entire process */
-static void exper_6_part2(struct proc **pp)
-{
- uint64_t begin = 0, diff = 0;
-
- if (*pp) {
- proc_incref(*pp, 1);
- spin_lock(&(*pp)->proc_lock);
- begin = start_timing();
- __proc_preempt_warnall(*pp, 1000000);
- spin_unlock(&(*pp)->proc_lock);
- spin_on((*pp)->procinfo->num_vcores > 1);
- diff = stop_timing(begin);
- printk("Took %llu usec (%llu nsec) to warn preempt all.\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- printk("[T]:006:%llu:%llu\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- }
}
-/* Experiment 7: preempt-raw, single core */
-static void exper_7_part2(struct proc **pp)
-{
- uint64_t begin = 0, diff = 0;
- bool self_ipi_pending = FALSE;
- uint32_t vcoreid, pcoreid = 7; // some core available on all systems
-
- if (*pp) {
- proc_incref(*pp, 1);
- spin_lock(&(*pp)->proc_lock);
- assert((*pp)->procinfo->pcoremap[pcoreid].valid);
- begin = start_timing();
- self_ipi_pending = __proc_preempt_core(*pp, pcoreid);
- spin_unlock(&(*pp)->proc_lock);
- __proc_kmsg_pending(*pp, self_ipi_pending);
- spin_on((*pp)->procinfo->pcoremap[pcoreid].valid);
- diff = stop_timing(begin);
- printk("Took %llu usec (%llu nsec) to raw-preempt one core.\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- printk("[T]:007:%llu:%llu\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- }
-}
-
-/* Experiment 8: preempt-warn, single core */
-static void exper_8_part2(struct proc **pp)
-{
- uint64_t begin = 0, diff = 0;
- uint32_t vcoreid, pcoreid = 7; // some core available on all systems
-
- if (*pp) {
- proc_incref(*pp, 1);
- spin_lock(&(*pp)->proc_lock);
- vcoreid = (*pp)->procinfo->pcoremap[pcoreid].vcoreid;
- assert((*pp)->procinfo->pcoremap[pcoreid].valid);
- begin = start_timing();
- __proc_preempt_warn(*pp, vcoreid, 1000000); // 1 sec
- spin_unlock(&(*pp)->proc_lock);
- spin_on((*pp)->procinfo->pcoremap[pcoreid].valid);
- diff = stop_timing(begin);
- printk("Took %llu usec (%llu nsec) to warn-preempt one core.\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- printk("[T]:008:%llu:%llu\n",
- diff * 1000000 / system_timing.tsc_freq,
- diff * 1000000000 / system_timing.tsc_freq);
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- }
-}
-
-/* Experiment 9: single notification time */
-static void exper_9_part2(struct proc **pp)
+void manager_waterman()
{
- struct notif_event ne = {0};
-
- if (*pp) {
- ne.ne_type = NE_ALARM;
- proc_incref(*pp, 1);
- printk("[T]:009:B:%llu\n", read_tsc());
- proc_notify(*pp, NE_ALARM, &ne);
- proc_destroy(*pp);
- proc_decref(*pp, 1);
- while (*pp) /* toggled in proc_free */
- cpu_relax();
- }
+ static bool first = true;
+ if (first)
+ mon_shell(0, 0, 0);
+ smp_idle();
+ assert(0);
}
-#endif /* __CONFIG_OSDI__ */
-
-#ifdef __sparc_v8__
-
-static char*
-itoa(int num, char* buf0, size_t base)
+void manager_yuzhu()
{
- if(base > 16)
- return NULL;
- char* buf = buf0;
- int len = 0, i;
-
- if(num < 0)
- {
- *buf++ = '-';
- num = -num;
- }
-
- do {
- buf[len++] = "0123456789abcdef"[num%base];
- num /= base;
- } while(num);
+ static uint8_t progress = 0;
+ static struct proc *p;
- for(i = 0; i < len/2; i++)
- {
- char temp = buf[i];
- buf[i] = buf[len-i-1];
- buf[len-i-1] = temp;
- }
- buf[len] = 0;
+ // for testing taking cores, check in case 1 for usage
+ uint32_t corelist[MAX_NUM_CORES];
+ uint32_t num = 3;
- return buf0;
-}
+ //create_server(init_num_cores, loop);
-void gsf_set_frame_cycles(int cycles)
-{
- store_alternate(26*4,2,cycles);
-}
+ monitor(0);
-void gsf_set_partition_credits(int partition, int credits)
-{
- store_alternate((32+partition)*4,2,credits);
-}
+ // quick_proc_run("hello", p);
-void gsf_set_core_partition(int core, int partition)
-{
- store_alternate((64+core)*4,2,partition);
}
-
-#endif
-
projects / akaros.git / blobdiff