[Here comes a rebased version against latest x86/mm]
In case "kgdbwait" is passed as kernel parameter, KGDB tries to set up
and connect to the front-end already during early_param evaluation. This
fails on x86 as the exception stack is not yet initialized, effectively
delaying kgdbwait until late-init.
Therefore, this patch hooks into the x86 trap initialization and
re-triggers the KGDB setup, including a potential early rendezvous with
gdb. As a precondition, KGDB's setup states are refactored once again to
allow multiple invocations of kgdb_early_entry and correct tracking of
pending kgdbwait requests.
Signed-off-by: Jan Kiszka <[EMAIL PROTECTED]>
---
arch/x86/kernel/traps_32.c | 4 +++
arch/x86/kernel/traps_64.c | 4 +++
include/linux/kgdb.h | 7 +++++-
kernel/kgdb.c | 52 +++++++++++++++++++++++++--------------------
4 files changed, 44 insertions(+), 23 deletions(-)
Index: b/arch/x86/kernel/traps_32.c
===================================================================
--- a/arch/x86/kernel/traps_32.c
+++ b/arch/x86/kernel/traps_32.c
@@ -25,6 +25,7 @@
#include <linux/utsname.h>
#include <linux/kprobes.h>
#include <linux/kexec.h>
+#include <linux/kgdb.h>
#include <linux/unwind.h>
#include <linux/uaccess.h>
#include <linux/nmi.h>
@@ -1215,6 +1216,9 @@ void __init trap_init(void)
*/
cpu_init();
+ /* With the TSS set up, it's now save to arm early KGDB. */
+ kgdb_early_entry();
+
trap_init_hook();
}
Index: b/arch/x86/kernel/traps_64.c
===================================================================
--- a/arch/x86/kernel/traps_64.c
+++ b/arch/x86/kernel/traps_64.c
@@ -27,6 +27,7 @@
#include <linux/nmi.h>
#include <linux/kprobes.h>
#include <linux/kexec.h>
+#include <linux/kgdb.h>
#include <linux/unwind.h>
#include <linux/uaccess.h>
#include <linux/bug.h>
@@ -1162,6 +1163,9 @@ void __init trap_init(void)
* Should be a barrier for any external CPU state.
*/
cpu_init();
+
+ /* With the TSS set up, it's now save to arm early KGDB. */
+ kgdb_early_entry();
}
Index: b/include/linux/kgdb.h
===================================================================
--- a/include/linux/kgdb.h
+++ b/include/linux/kgdb.h
@@ -43,7 +43,8 @@ extern struct task_struct *kgdb_contthre
enum kgdb_initstate {
KGDB_UNINITIALIZED = 0,
- KGDB_SEMI_INITIALIZED,
+ KGDB_ARCH_INITIALIZED,
+ KGDB_DELAYED_CONNECTION,
KGDB_FULLY_INITIALIZED
};
@@ -287,6 +288,8 @@ int kgdb_handle_exception(int ex_vector,
struct pt_regs *regs);
int kgdb_nmihook(int cpu, void *regs);
+void __init kgdb_early_entry(void);
+
extern int debugger_step;
extern atomic_t debugger_active;
@@ -296,6 +299,8 @@ extern atomic_t debugger_active;
#else /* !CONFIG_KGDB */
static const atomic_t debugger_active = ATOMIC_INIT(0);
+
+static inline void kgdb_early_entry(void) { }
#endif /* !CONFIG_KGDB */
#endif /* _KGDB_H_ */
Index: b/kernel/kgdb.c
===================================================================
--- a/kernel/kgdb.c
+++ b/kernel/kgdb.c
@@ -2104,33 +2104,41 @@ void kgdb_unregister_io_module(struct kg
}
EXPORT_SYMBOL_GPL(kgdb_unregister_io_module);
+static void __init kgdb_initial_breakpoint(void)
+{
+ printk(KERN_CRIT "kgdb: Waiting for connection from remote gdb...\n");
+ breakpoint();
+}
+
/*
* This function can be called very early, either via early_param() or
* an explicit breakpoint() early on.
*/
-static void __init kgdb_early_entry(void)
+void __init kgdb_early_entry(void)
{
+ int need_break = (kgdb_state == KGDB_DELAYED_CONNECTION);
+
/* Let the architecture do any setup that it needs to. */
- kgdb_arch_init();
+ if (kgdb_state == KGDB_UNINITIALIZED) {
+ kgdb_arch_init();
+ kgdb_state = KGDB_ARCH_INITIALIZED;
+ }
/*
* Don't try and do anything until the architecture is able to
* setup the exception stack. In this case, it is up to the
* architecture to hook in and look at us when they are ready.
*/
- if (!EXCEPTION_STACK_READY()) {
- kgdb_state = KGDB_SEMI_INITIALIZED;
- /* any kind of break point is deferred to late_init */
+ if (!EXCEPTION_STACK_READY())
return;
- }
/*
* Now try the I/O.
* For early entry kgdb_io_ops.init must be defined
*/
if (!kgdb_io_ops.init || kgdb_io_ops.init()) {
- /* Try again later. */
- kgdb_state = KGDB_SEMI_INITIALIZED;
+ printk(KERN_ERR "kgdb: Could not setup core I/O for KGDB.\n");
+ printk(KERN_INFO "kgdb: Defering I/O setup to late init.\n");
return;
}
@@ -2145,6 +2153,9 @@ static void __init kgdb_early_entry(void
*/
if (kgdb_io_ops.init)
kgdb_register_for_panic();
+
+ if (need_break)
+ kgdb_initial_breakpoint();
}
/*
@@ -2155,14 +2166,16 @@ static void __init kgdb_early_entry(void
*/
static int __init kgdb_late_entry(void)
{
- int need_break = (kgdb_state == KGDB_SEMI_INITIALIZED);
+ int need_break = (kgdb_state == KGDB_DELAYED_CONNECTION);
/*
* If we haven't tried to initialize KGDB yet, we need to call
* kgdb_arch_init before moving onto the I/O.
*/
- if (kgdb_state == KGDB_UNINITIALIZED)
+ if (kgdb_state == KGDB_UNINITIALIZED) {
kgdb_arch_init();
+ kgdb_state = KGDB_ARCH_INITIALIZED;
+ }
if (kgdb_state != KGDB_FULLY_INITIALIZED) {
if (kgdb_io_ops.init && kgdb_io_ops.init()) {
@@ -2177,6 +2190,7 @@ static int __init kgdb_late_entry(void)
printk(KERN_INFO "kgdb: Defering I/O setup to kernel "
"module.\n");
memset(&kgdb_io_ops, 0, sizeof(struct kgdb_io));
+ need_break = 0;
}
kgdb_internal_init();
@@ -2198,11 +2212,8 @@ static int __init kgdb_late_entry(void)
if (kgdb_io_ops.late_init)
kgdb_io_ops.late_init();
- if (need_break) {
- printk(KERN_CRIT "kgdb: Waiting for connection from remote"
- " gdb...\n");
- breakpoint();
- }
+ if (need_break)
+ kgdb_initial_breakpoint();
return 0;
}
@@ -2290,14 +2301,11 @@ static int __init opt_kgdb_enter(char *s
kgdb_early_entry();
attachwait = 1;
- if (kgdb_state == KGDB_FULLY_INITIALIZED)
- printk(KERN_CRIT "Waiting for connection from remote gdb...\n");
- else {
- printk(KERN_CRIT "KGDB cannot initialize I/O yet.\n");
- return 0;
- }
- breakpoint();
+ if (kgdb_state == KGDB_FULLY_INITIALIZED)
+ kgdb_initial_breakpoint();
+ else
+ kgdb_state = KGDB_DELAYED_CONNECTION;
return 0;
}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
