LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP.
@ 2015-02-26 3:57 Wang Nan
2015-02-26 4:10 ` Steven Rostedt
2015-02-26 4:48 ` Masami Hiramatsu
0 siblings, 2 replies; 3+ messages in thread
From: Wang Nan @ 2015-02-26 3:57 UTC (permalink / raw)
To: mingo, hpa, tglx, x86, luto, oleg, masami.hiramatsu.pt,
dave.hansen, rostedt
Cc: linux-kernel, lizefan
Before this patch early_trap_init() installs DEBUG_STACK for X86_TRAP_BP
and X86_TRAP_DB. However, DEBUG_STACK doesn't work correctly until
cpu_init() <-- trap_init().
This patch passes 0 to set_intr_gate_ist() and
set_system_intr_gate_ist() instead of DEBUG_STACK to let it use same
stack as kernel, and installs DEBUG_STACK for them in trap_init().
As core runs at ring 0 between early_trap_init() and trap_init(), there
is no chance to get a bad stack before trap_init().
As NMI is also enabled in trap_init(), we don't need to care about
is_debug_stack() and related things used in arch/x86/kernel/nmi.c.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
---
arch/x86/kernel/traps.c | 20 ++++++++++++++++++--
1 file changed, 18 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
index 9d2073e..a9b8640 100644
--- a/arch/x86/kernel/traps.c
+++ b/arch/x86/kernel/traps.c
@@ -925,9 +925,16 @@ dotraplinkage void do_iret_error(struct pt_regs *regs, long error_code)
/* Set of traps needed for early debugging. */
void __init early_trap_init(void)
{
- set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
+ /*
+ * Don't set ist to DEBUG_STACK as it doesn't work until TSS is
+ * ready in cpu_init() <-- trap_init(). Before trap_init(), CPU
+ * runs at ring 0 so there should be impossible to hit a invalid
+ * stack. Use original stack is enough. DEBUG_STACK will be
+ * equipped after cpu_init() in trap_init().
+ */
+ set_intr_gate_ist(X86_TRAP_DB, &debug, 0);
/* int3 can be called from all */
- set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
+ set_system_intr_gate_ist(X86_TRAP_BP, &int3, 0);
#ifdef CONFIG_X86_32
set_intr_gate(X86_TRAP_PF, page_fault);
#endif
@@ -1005,6 +1012,15 @@ void __init trap_init(void)
*/
cpu_init();
+ /*
+ * X86_TRAP_DB and X86_TRAP_BP have been setup
+ * in early_trap_init(). However, DEBUG_STACK works only after
+ * cpu_init() load TSS. See comments in early_trap_init().
+ */
+ set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
+ /* int3 can be called from all */
+ set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
+
x86_init.irqs.trap_init();
#ifdef CONFIG_X86_64
--
1.8.4
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP.
2015-02-26 3:57 [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP Wang Nan
@ 2015-02-26 4:10 ` Steven Rostedt
2015-02-26 4:48 ` Masami Hiramatsu
1 sibling, 0 replies; 3+ messages in thread
From: Steven Rostedt @ 2015-02-26 4:10 UTC (permalink / raw)
To: Wang Nan
Cc: mingo, hpa, tglx, x86, luto, oleg, masami.hiramatsu.pt,
dave.hansen, linux-kernel, lizefan
On Thu, 26 Feb 2015 11:57:56 +0800
Wang Nan <wangnan0@huawei.com> wrote:
> Before this patch early_trap_init() installs DEBUG_STACK for X86_TRAP_BP
> and X86_TRAP_DB. However, DEBUG_STACK doesn't work correctly until
> cpu_init() <-- trap_init().
>
> This patch passes 0 to set_intr_gate_ist() and
> set_system_intr_gate_ist() instead of DEBUG_STACK to let it use same
> stack as kernel, and installs DEBUG_STACK for them in trap_init().
>
> As core runs at ring 0 between early_trap_init() and trap_init(), there
> is no chance to get a bad stack before trap_init().
>
> As NMI is also enabled in trap_init(), we don't need to care about
> is_debug_stack() and related things used in arch/x86/kernel/nmi.c.
Looks fine to me, except for some grammar issues in the comments.
>
> Signed-off-by: Wang Nan <wangnan0@huawei.com>
> ---
> arch/x86/kernel/traps.c | 20 ++++++++++++++++++--
> 1 file changed, 18 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
> index 9d2073e..a9b8640 100644
> --- a/arch/x86/kernel/traps.c
> +++ b/arch/x86/kernel/traps.c
> @@ -925,9 +925,16 @@ dotraplinkage void do_iret_error(struct pt_regs *regs, long error_code)
> /* Set of traps needed for early debugging. */
> void __init early_trap_init(void)
> {
> - set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
> + /*
> + * Don't set ist to DEBUG_STACK as it doesn't work until TSS is
> + * ready in cpu_init() <-- trap_init(). Before trap_init(), CPU
> + * runs at ring 0 so there should be impossible to hit a invalid
"so it should be impossible", although I might prefer, "it is
impossible to", but nothing is ever guaranteed.
"an invalid"
> + * stack. Use original stack is enough. DEBUG_STACK will be
"Using the original stack works well enough at this early stage."
> + * equipped after cpu_init() in trap_init().
> + */
> + set_intr_gate_ist(X86_TRAP_DB, &debug, 0);
> /* int3 can be called from all */
> - set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
> + set_system_intr_gate_ist(X86_TRAP_BP, &int3, 0);
> #ifdef CONFIG_X86_32
> set_intr_gate(X86_TRAP_PF, page_fault);
> #endif
> @@ -1005,6 +1012,15 @@ void __init trap_init(void)
> */
> cpu_init();
>
> + /*
> + * X86_TRAP_DB and X86_TRAP_BP have been setup
> + * in early_trap_init(). However, DEBUG_STACK works only after
> + * cpu_init() load TSS. See comments in early_trap_init().
"cpu_init() loads TSS"
-- Steve
> + */
> + set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
> + /* int3 can be called from all */
> + set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
> +
> x86_init.irqs.trap_init();
>
> #ifdef CONFIG_X86_64
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP.
2015-02-26 3:57 [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP Wang Nan
2015-02-26 4:10 ` Steven Rostedt
@ 2015-02-26 4:48 ` Masami Hiramatsu
1 sibling, 0 replies; 3+ messages in thread
From: Masami Hiramatsu @ 2015-02-26 4:48 UTC (permalink / raw)
To: Wang Nan
Cc: mingo, hpa, tglx, x86, luto, oleg, dave.hansen, rostedt,
linux-kernel, lizefan
(2015/02/26 12:57), Wang Nan wrote:
> Before this patch early_trap_init() installs DEBUG_STACK for X86_TRAP_BP
> and X86_TRAP_DB. However, DEBUG_STACK doesn't work correctly until
> cpu_init() <-- trap_init().
>
> This patch passes 0 to set_intr_gate_ist() and
> set_system_intr_gate_ist() instead of DEBUG_STACK to let it use same
> stack as kernel, and installs DEBUG_STACK for them in trap_init().
>
> As core runs at ring 0 between early_trap_init() and trap_init(), there
> is no chance to get a bad stack before trap_init().
Thanks for finding the problem on it! :)
Agreed, until initializing DEBUG_STACK, it should not be used.
>
> As NMI is also enabled in trap_init(), we don't need to care about
> is_debug_stack() and related things used in arch/x86/kernel/nmi.c.
Looks good to me, at least its code side. Please fix the comment according
to Steven's suggestion.
And feel free to add my reviewed-by for this.
Reviewed-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Thank you,
>
> Signed-off-by: Wang Nan <wangnan0@huawei.com>
> ---
> arch/x86/kernel/traps.c | 20 ++++++++++++++++++--
> 1 file changed, 18 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
> index 9d2073e..a9b8640 100644
> --- a/arch/x86/kernel/traps.c
> +++ b/arch/x86/kernel/traps.c
> @@ -925,9 +925,16 @@ dotraplinkage void do_iret_error(struct pt_regs *regs, long error_code)
> /* Set of traps needed for early debugging. */
> void __init early_trap_init(void)
> {
> - set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
> + /*
> + * Don't set ist to DEBUG_STACK as it doesn't work until TSS is
> + * ready in cpu_init() <-- trap_init(). Before trap_init(), CPU
> + * runs at ring 0 so there should be impossible to hit a invalid
> + * stack. Use original stack is enough. DEBUG_STACK will be
> + * equipped after cpu_init() in trap_init().
> + */
> + set_intr_gate_ist(X86_TRAP_DB, &debug, 0);
> /* int3 can be called from all */
> - set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
> + set_system_intr_gate_ist(X86_TRAP_BP, &int3, 0);
> #ifdef CONFIG_X86_32
> set_intr_gate(X86_TRAP_PF, page_fault);
> #endif
> @@ -1005,6 +1012,15 @@ void __init trap_init(void)
> */
> cpu_init();
>
> + /*
> + * X86_TRAP_DB and X86_TRAP_BP have been setup
> + * in early_trap_init(). However, DEBUG_STACK works only after
> + * cpu_init() load TSS. See comments in early_trap_init().
> + */
> + set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK);
> + /* int3 can be called from all */
> + set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK);
> +
> x86_init.irqs.trap_init();
>
> #ifdef CONFIG_X86_64
>
--
Masami HIRAMATSU
Software Platform Research Dept. Linux Technology Research Center
Hitachi, Ltd., Yokohama Research Laboratory
E-mail: masami.hiramatsu.pt@hitachi.com
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2015-02-26 4:48 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-02-26 3:57 [PATCH] x86, traps: Enable DEBUG_STACK after cpu_init() for TRAP_DB/BP Wang Nan
2015-02-26 4:10 ` Steven Rostedt
2015-02-26 4:48 ` Masami Hiramatsu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).