Skip to contentSkip to conversation

x86_64: use kstack_end() in dumpstack_64.c

PATCH Applied Adrien Schildknecht 4 participants 7 messages Feb 22, 2015 – Mar 3, 2015 x86

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86_64: use kstack_end() in dumpstack_64.c

i386 is already using kstack_end() in dumpstack_32.c. We should also use it to make the code clearer and unify the stack printing logic some more.

This patch depends on patch "x86: fix output of show_stack_log_lvl()"

---
 arch/x86/kernel/dumpstack_64.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/dumpstack_64.c +1 -1
index 553573b..5f1c626 100644
@@ -280,7 +280,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
 				pr_cont(" <EOI> ");
 			}
 		} else {
-		if (((long) stack & (THREAD_SIZE-1)) == 0)
+		if (kstack_end(stack))
 			break;
 		}
 		if ((i % STACKSLOTS_PER_LINE) == 0) {
Signature · 1 line
2.2.1
#1 lore raw
Quoted 3 lines · On Sun, 22 Feb 2015 16:23:58 +0100 Adrien Schildknecht…

i386 is already using kstack_end() in dumpstack_32.c. We should also use it to make the code clearer and unify the stack printing logic some more.

Looks fine to me.

Quoted 2 lines

This patch depends on patch "x86: fix output of show_stack_log_lvl()"

I'm curious to what the dependency is? What would break if we apply this without that patch?

-- Steve

Quoted 19 lines
---
 arch/x86/kernel/dumpstack_64.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/dumpstack_64.c +1 -1
index 553573b..5f1c626 100644
@@ -280,7 +280,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
 				pr_cont(" <EOI> ");
 			}
 		} else {
-		if (((long) stack & (THREAD_SIZE-1)) == 0)
+		if (kstack_end(stack))
 			break;
 		}
 		if ((i % STACKSLOTS_PER_LINE) == 0) {
maintainer
#2 lore raw
Quoted 2 lines · Steven Rostedt

I'm curious to what the dependency is? What would break if we apply this without that patch?

Nah, no dependency - it causes just a little fuzz when the previous one is missing.

I'll apply it to the proper tree and fixup the commit message :)

Signature · 5 lines
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--
#3 lore raw
In reply to Steven Rostedt
Quoted 5 lines
Quoted 2 lines

This patch depends on patch "x86: fix output of
show_stack_log_lvl()"

I'm curious to what the dependency is? What would break if we apply this without that patch?

Maybe I misunderstood "dependency".
I mean that this patch won't apply properly since a line of the context was changed by the previous patch.
"if ((i % STACKSLOTS_PER_LINE) == 0) {"

Signature · 1 line
Adrien Schildknecht
#4 lore raw
Quoted 11 lines · On Mon, 23 Feb 2015 18:29:05 +0100 Adrien Schildknecht…
Quoted 5 lines
Quoted 2 lines

This patch depends on patch "x86: fix output of
show_stack_log_lvl()"

I'm curious to what the dependency is? What would break if we apply this without that patch?

Maybe I misunderstood "dependency".
I mean that this patch won't apply properly since a line of the context was changed by the previous patch.
"if ((i % STACKSLOTS_PER_LINE) == 0) {"

Yeah, "dependency" is a bit ambiguous here. It can mean both "this patch wont work without said patch" as well as "this patch wont apply without said patch". Probably would have been better to say that it won't apply to avoid the ambiguity.

Also, these types of notes are best stated after the '---' line. As it's good information for the maintainer pulling in the patch, but should not go in the change log. Same for comments like "Changes since version 1 of this series".

That is:

i386 is already using kstack_end() in dumpstack_32.c. We should also use it to make the code clearer and unify the stack printing logic some more.

---
Note: This patch depends on patch "x86: fix output of show_stack_log_lvl()" or it might not apply properly.

 arch/x86/kernel/dumpstack_64.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

The change log for the above stops at the '---'.

-- Steve

maintainer Applied
#5 lore raw
In reply to Adrien Schildknecht
Quoted 7 lines · Adrien Schildknecht

i386 is already using kstack_end() in dumpstack_32.c. We should also use it to make the code clearer and unify the stack printing logic some more.

This patch depends on patch "x86: fix output of show_stack_log_lvl()"

Applied, thanks.

Signature · 5 lines
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--
#6 lore raw

[tip:x86/debug] x86/kernel: Use kstack_end() in dumpstack_64.c

In reply to Adrien Schildknecht
x86/kernel: Use kstack_end() in dumpstack_64.c
Commit-ID:  04769ae3ac72f86324a189b69f53bf3bfb61acfd
Gitweb:     http://git.kernel.org/tip/04769ae3ac72f86324a189b69f53bf3bfb61acfd
Author:     Adrien Schildknecht <[email protected]>
AuthorDate: Sun, 22 Feb 2015 16:23:58 +0100
Committer:  Borislav Petkov <[email protected]>
CommitDate: Mon, 23 Feb 2015 18:37:13 +0100

x86/kernel: Use kstack_end() in dumpstack_64.c

i386 is already using kstack_end() in dumpstack_32.c. We should also use it to make the code clearer and unify the stack printing logic some more.

---
 arch/x86/kernel/dumpstack_64.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/dumpstack_64.c +1 -1
index 553573b..5f1c626 100644
@@ -280,7 +280,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
 				pr_cont(" <EOI> ");
 			}
 		} else {
-		if (((long) stack & (THREAD_SIZE-1)) == 0)
+		if (kstack_end(stack))
 			break;
 		}
 		if ((i % STACKSLOTS_PER_LINE) == 0) {