Skip to contentSkip to conversation

ftrace: updates for powerpc

PATCH In review Steven Rostedt 5 participants 15 messages Feb 6, 2009 – May 11, 2010

Subsystems (2)
lore
Share & tools
lore

Steven Rostedt

author maintainer
#0 lore raw

cover ftrace: updates for powerpc

Paul,

The first patch is a ftrace clean up to use the proper pr_debug macro.

The second patch is to fix a compile error when dynamic ftrace is compiled in, but modules are not.

The following patches are in:

git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-2.6-trace.git

branch: ppc/ftrace

Steven Rostedt (2):
ftrace, powerpc: replace debug macro with proper pr_deug powerpc, ftrace: fix compile error when modules not configured

----

 arch/powerpc/kernel/ftrace.c |   42 +++++++++++++++++++++++++-----------------
 1 files changed, 25 insertions(+), 17 deletions(-)
author maintainer
#1 lore raw

PATCH 1/2 ftrace, powerpc: replace debug macro with proper pr_deug

From: Steven Rostedt <[email protected]>

Impact: cleanup

The PowerPC ftrace code uses a hacked up DEBUGP macro for prints. This patch converts it to the standard pr_debug.

---
 arch/powerpc/kernel/ftrace.c |   26 ++++++++++----------------
 1 files changed, 10 insertions(+), 16 deletions(-)
arch/powerpc/kernel/ftrace.c +10 -16
index 5355244..a913f91 100644
@@ -20,12 +20,6 @@
 #include <asm/code-patching.h>
 #include <asm/ftrace.h>
 
-#if 0
-#define DEBUGP printk
-#else
-#define DEBUGP(fmt , ...)	do { } while (0)
-#endif
-
 static unsigned int ftrace_nop = PPC_NOP_INSTR;
 
 #ifdef CONFIG_PPC32
@@ -175,7 +169,7 @@ __ftrace_make_nop(struct module *mod,
 	 * 0xe8, 0x4c, 0x00, 0x28,    ld      r2,40(r12)
 	 */
 
-	DEBUGP("ip:%lx jumps to %lx r2: %lx", ip, tramp, mod->arch.toc);
+	pr_debug("ip:%lx jumps to %lx r2: %lx", ip, tramp, mod->arch.toc);
 
 	/* Find where the trampoline jumps to */
 	if (probe_kernel_read(jmp, (void *)tramp, sizeof(jmp))) {
@@ -183,7 +177,7 @@ __ftrace_make_nop(struct module *mod,
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x", jmp[0], jmp[1]);
 
 	/* verify that this is what we expect it to be */
 	if (((jmp[0] & 0xffff0000) != 0x3d820000) ||
@@ -198,18 +192,18 @@ __ftrace_make_nop(struct module *mod,
 	offset = (unsigned)((unsigned short)jmp[0]) << 16 |
 		(unsigned)((unsigned short)jmp[1]);
 
-	DEBUGP(" %x ", offset);
+	pr_debug(" %x ", offset);
 
 	/* get the address this jumps too */
 	tramp = mod->arch.toc + offset + 32;
-	DEBUGP("toc: %lx", tramp);
+	pr_debug("toc: %lx", tramp);
 
 	if (probe_kernel_read(jmp, (void *)tramp, 8)) {
 		printk(KERN_ERR "Failed to read %lx\n", tramp);
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x\n", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x\n", jmp[0], jmp[1]);
 
 	ptr = ((unsigned long)jmp[0] << 32) + jmp[1];
 
@@ -286,7 +280,7 @@ __ftrace_make_nop(struct module *mod,
 	 *  0x4e, 0x80, 0x04, 0x20  bctr
 	 */
 
-	DEBUGP("ip:%lx jumps to %lx", ip, tramp);
+	pr_debug("ip:%lx jumps to %lx", ip, tramp);
 
 	/* Find where the trampoline jumps to */
 	if (probe_kernel_read(jmp, (void *)tramp, sizeof(jmp))) {
@@ -294,7 +288,7 @@ __ftrace_make_nop(struct module *mod,
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x ", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x ", jmp[0], jmp[1]);
 
 	/* verify that this is what we expect it to be */
 	if (((jmp[0] & 0xffff0000) != 0x3d600000) ||
@@ -310,7 +304,7 @@ __ftrace_make_nop(struct module *mod,
 	if (tramp & 0x8000)
 		tramp -= 0x10000;
 
-	DEBUGP(" %x ", tramp);
+	pr_debug(" %x ", tramp);
 
 	if (tramp != addr) {
 		printk(KERN_ERR
@@ -413,7 +407,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	/* ld r2,40(r1) */
 	op[1] = 0xe8410028;
 
-	DEBUGP("write to %lx\n", rec->ip);
+	pr_debug("write to %lx\n", rec->ip);
 
 	if (probe_kernel_write((void *)ip, op, MCOUNT_INSN_SIZE * 2))
 		return -EPERM;
@@ -453,7 +447,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 		return -EINVAL;
 	}
 
-	DEBUGP("write to %lx\n", rec->ip);
+	pr_debug("write to %lx\n", rec->ip);
 
 	if (probe_kernel_write((void *)ip, &op, MCOUNT_INSN_SIZE))
 		return -EPERM;
Signature · 3 lines
1.5.6.5

-- 
author maintainer
#2 lore raw

Re: [PATCH 1/2] ftrace, powerpc: replace debug macro with proper pr_deug

[ forgot to add linuxppc ]

Quoted 120 lines · Steven Rostedt

From: Steven Rostedt <[email protected]>

Impact: cleanup

The PowerPC ftrace code uses a hacked up DEBUGP macro for prints. This patch converts it to the standard pr_debug.

---
 arch/powerpc/kernel/ftrace.c |   26 ++++++++++----------------
 1 files changed, 10 insertions(+), 16 deletions(-)
arch/powerpc/kernel/ftrace.c +10 -16
index 5355244..a913f91 100644
@@ -20,12 +20,6 @@
 #include <asm/code-patching.h>
 #include <asm/ftrace.h>
 
-#if 0
-#define DEBUGP printk
-#else
-#define DEBUGP(fmt , ...)	do { } while (0)
-#endif
-
 static unsigned int ftrace_nop = PPC_NOP_INSTR;
 
 #ifdef CONFIG_PPC32
@@ -175,7 +169,7 @@ __ftrace_make_nop(struct module *mod,
 	 * 0xe8, 0x4c, 0x00, 0x28,    ld      r2,40(r12)
 	 */
 
-	DEBUGP("ip:%lx jumps to %lx r2: %lx", ip, tramp, mod->arch.toc);
+	pr_debug("ip:%lx jumps to %lx r2: %lx", ip, tramp, mod->arch.toc);
 
 	/* Find where the trampoline jumps to */
 	if (probe_kernel_read(jmp, (void *)tramp, sizeof(jmp))) {
@@ -183,7 +177,7 @@ __ftrace_make_nop(struct module *mod,
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x", jmp[0], jmp[1]);
 
 	/* verify that this is what we expect it to be */
 	if (((jmp[0] & 0xffff0000) != 0x3d820000) ||
@@ -198,18 +192,18 @@ __ftrace_make_nop(struct module *mod,
 	offset = (unsigned)((unsigned short)jmp[0]) << 16 |
 		(unsigned)((unsigned short)jmp[1]);
 
-	DEBUGP(" %x ", offset);
+	pr_debug(" %x ", offset);
 
 	/* get the address this jumps too */
 	tramp = mod->arch.toc + offset + 32;
-	DEBUGP("toc: %lx", tramp);
+	pr_debug("toc: %lx", tramp);
 
 	if (probe_kernel_read(jmp, (void *)tramp, 8)) {
 		printk(KERN_ERR "Failed to read %lx\n", tramp);
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x\n", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x\n", jmp[0], jmp[1]);
 
 	ptr = ((unsigned long)jmp[0] << 32) + jmp[1];
 
@@ -286,7 +280,7 @@ __ftrace_make_nop(struct module *mod,
 	 *  0x4e, 0x80, 0x04, 0x20  bctr
 	 */
 
-	DEBUGP("ip:%lx jumps to %lx", ip, tramp);
+	pr_debug("ip:%lx jumps to %lx", ip, tramp);
 
 	/* Find where the trampoline jumps to */
 	if (probe_kernel_read(jmp, (void *)tramp, sizeof(jmp))) {
@@ -294,7 +288,7 @@ __ftrace_make_nop(struct module *mod,
 		return -EFAULT;
 	}
 
-	DEBUGP(" %08x %08x ", jmp[0], jmp[1]);
+	pr_debug(" %08x %08x ", jmp[0], jmp[1]);
 
 	/* verify that this is what we expect it to be */
 	if (((jmp[0] & 0xffff0000) != 0x3d600000) ||
@@ -310,7 +304,7 @@ __ftrace_make_nop(struct module *mod,
 	if (tramp & 0x8000)
 		tramp -= 0x10000;
 
-	DEBUGP(" %x ", tramp);
+	pr_debug(" %x ", tramp);
 
 	if (tramp != addr) {
 		printk(KERN_ERR
@@ -413,7 +407,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	/* ld r2,40(r1) */
 	op[1] = 0xe8410028;
 
-	DEBUGP("write to %lx\n", rec->ip);
+	pr_debug("write to %lx\n", rec->ip);
 
 	if (probe_kernel_write((void *)ip, op, MCOUNT_INSN_SIZE * 2))
 		return -EPERM;
@@ -453,7 +447,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 		return -EINVAL;
 	}
 
-	DEBUGP("write to %lx\n", rec->ip);
+	pr_debug("write to %lx\n", rec->ip);
 
 	if (probe_kernel_write((void *)ip, &op, MCOUNT_INSN_SIZE))
 		return -EPERM;
Signature · 3 lines
1.5.6.5

-- 
Acked-by on v1 1/2
#3 lore raw

Re: [PATCH 1/2] ftrace, powerpc: replace debug macro with proper pr_deug

In reply to Steven Rostedt
Quoted 10 lines · Steven Rostedt

plain text document attachment (0001-ftrace-powerpc-replace-debug-macro-with-proper-pr_.patch)
From: Steven Rostedt <[email protected]>

Impact: cleanup

The PowerPC ftrace code uses a hacked up DEBUGP macro for prints. This patch converts it to the standard pr_debug.

---

Want me to go through powerpc tree or you'll merge via Ingo ?

author maintainer
#4 lore raw

Re: [PATCH 1/2] ftrace, powerpc: replace debug macro with proper pr_deug

Quoted 10 lines · Benjamin Herrenschmidt
Quoted 9 lines · Steven Rostedt

plain text document attachment (0001-ftrace-powerpc-replace-debug-macro-with-proper-pr_.patch)
From: Steven Rostedt <[email protected]>

Impact: cleanup

The PowerPC ftrace code uses a hacked up DEBUGP macro for prints. This patch converts it to the standard pr_debug.

Quoted 6 lines
Quoted 1 line

---

Want me to go through powerpc tree or you'll merge via Ingo ?

These two changes are powerpc specific. I think it may get more testing going through the powerpc tree than through tip.

-- Steve

author maintainer
#5 lore raw

PATCH 2/2 powerpc, ftrace: fix compile error when modules not configured

In reply to Steven Rostedt

From: Steven Rostedt <[email protected]>

Michael Neuling reported a compile bug when dynamic ftrace was configured in and modules were not. This was due to the ftrace code referencing module specific structures.

---
 arch/powerpc/kernel/ftrace.c |   16 +++++++++++++++-
 1 files changed, 15 insertions(+), 1 deletions(-)
arch/powerpc/kernel/ftrace.c +15 -1
index a913f91..88c641d 100644
@@ -113,6 +113,8 @@ static int test_24bit_addr(unsigned long ip, unsigned long addr)
 	return create_branch((unsigned int *)ip, addr, 0);
 }
 
+#ifdef CONFIG_MODULES
+
 static int is_bl_op(unsigned int op)
 {
 	return (op & 0xfc000003) == 0x48000001;
@@ -323,6 +325,7 @@ __ftrace_make_nop(struct module *mod,
 	return 0;
 }
 #endif /* PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_nop(struct module *mod,
 		    struct dyn_ftrace *rec, unsigned long addr)
@@ -342,6 +345,7 @@ int ftrace_make_nop(struct module *mod,
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * We should either already have a pointer to the module
@@ -366,9 +370,13 @@ int ftrace_make_nop(struct module *mod,
 		mod = rec->arch.mod;
 
 	return __ftrace_make_nop(mod, rec, addr);
-
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
+#ifdef CONFIG_MODULES
 #ifdef CONFIG_PPC64
 static int
 __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
@@ -457,6 +465,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	return 0;
 }
 #endif /* CONFIG_PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 {
@@ -475,6 +484,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * Being that we are converting from nop, it had better
@@ -486,6 +496,10 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	}
 
 	return __ftrace_make_call(rec, addr);
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
 int ftrace_update_ftrace_func(ftrace_func_t func)
Signature · 3 lines
1.5.6.5

-- 
author maintainer
#6 lore raw

Re: [PATCH 2/2] powerpc, ftrace: fix compile error when modules not configured

[ forgot to add linuxppc ]

Quoted 89 lines · Steven Rostedt

From: Steven Rostedt <[email protected]>

Michael Neuling reported a compile bug when dynamic ftrace was configured in and modules were not. This was due to the ftrace code referencing module specific structures.

---
 arch/powerpc/kernel/ftrace.c |   16 +++++++++++++++-
 1 files changed, 15 insertions(+), 1 deletions(-)
arch/powerpc/kernel/ftrace.c +15 -1
index a913f91..88c641d 100644
@@ -113,6 +113,8 @@ static int test_24bit_addr(unsigned long ip, unsigned long addr)
 	return create_branch((unsigned int *)ip, addr, 0);
 }
 
+#ifdef CONFIG_MODULES
+
 static int is_bl_op(unsigned int op)
 {
 	return (op & 0xfc000003) == 0x48000001;
@@ -323,6 +325,7 @@ __ftrace_make_nop(struct module *mod,
 	return 0;
 }
 #endif /* PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_nop(struct module *mod,
 		    struct dyn_ftrace *rec, unsigned long addr)
@@ -342,6 +345,7 @@ int ftrace_make_nop(struct module *mod,
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * We should either already have a pointer to the module
@@ -366,9 +370,13 @@ int ftrace_make_nop(struct module *mod,
 		mod = rec->arch.mod;
 
 	return __ftrace_make_nop(mod, rec, addr);
-
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
+#ifdef CONFIG_MODULES
 #ifdef CONFIG_PPC64
 static int
 __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
@@ -457,6 +465,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	return 0;
 }
 #endif /* CONFIG_PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 {
@@ -475,6 +484,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * Being that we are converting from nop, it had better
@@ -486,6 +496,10 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	}
 
 	return __ftrace_make_call(rec, addr);
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
 int ftrace_update_ftrace_func(ftrace_func_t func)
Signature · 3 lines
1.5.6.5

-- 
#7 lore raw

Re: [PATCH 2/2] powerpc, ftrace: fix compile error when modules not configured

In reply to Steven Rostedt
Quoted 8 lines · In message <[email protected]> you wrote:

From: Steven Rostedt <[email protected]>

Michael Neuling reported a compile bug when dynamic ftrace was configured in and modules were not. This was due to the ftrace code referencing module specific structures.

Steve,

Thanks, fixes the error I was seeing.

As an aside, is there anyway we can merge some of the code in arch/powerpc/kernel/ftrace.c between 32 and 64bit? There seems to be a lot of repeated code in there with only minor changes.

Mikey

arch/powerpc/kernel/ftrace.c static int test_24bit_addr(unsigned long ip, unsigned lon

Quoted 9 lines
---
 arch/powerpc/kernel/ftrace.c |   16 +++++++++++++++-
 1 files changed, 15 insertions(+), 1 deletions(-)
arch/powerpc/kernel/ftrace.c +0 -0
index a913f91..88c641d 100644
@@ -113,6 +113,8 @@ static int test_24bit_addr(unsigned long ip, unsigned lon

g addr)

arch/powerpc/kernel/ftrace.c __ftrace_make_call(struct dyn_ftrace *rec, unsigned long

Quoted 40 lines
arch/powerpc/kernel/ftrace.c +9 -1
 	return create_branch((unsigned int *)ip, addr, 0);
 }
 
+#ifdef CONFIG_MODULES
+
 static int is_bl_op(unsigned int op)
 {
 	return (op & 0xfc000003) == 0x48000001;
@@ -323,6 +325,7 @@ __ftrace_make_nop(struct module *mod,
 	return 0;
 }
 #endif /* PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_nop(struct module *mod,
 		    struct dyn_ftrace *rec, unsigned long addr)
@@ -342,6 +345,7 @@ int ftrace_make_nop(struct module *mod,
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * We should either already have a pointer to the module
@@ -366,9 +370,13 @@ int ftrace_make_nop(struct module *mod,
 		mod = rec->arch.mod;
 
 	return __ftrace_make_nop(mod, rec, addr);
-
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
+#ifdef CONFIG_MODULES
 #ifdef CONFIG_PPC64
 static int
 __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
@@ -457,6 +465,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long 

addr)

arch/powerpc/kernel/ftrace.c int ftrace_make_call(struct dyn_ftrace *rec, unsigned lon

Quoted 8 lines
arch/powerpc/kernel/ftrace.c +1 -0
 	return 0;
 }
 #endif /* CONFIG_PPC64 */
+#endif /* CONFIG_MODULES */
 
 int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 {
@@ -475,6 +484,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned lon

g addr)

arch/powerpc/kernel/ftrace.c int ftrace_make_call(struct dyn_ftrace *rec, unsigned lo

Quoted 8 lines
arch/powerpc/kernel/ftrace.c +1 -0
 		return ftrace_modify_code(ip, old, new);
 	}
 
+#ifdef CONFIG_MODULES
 	/*
 	 * Out of range jumps are called from modules.
 	 * Being that we are converting from nop, it had better
@@ -486,6 +496,10 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned lo

ng addr)

Quoted 15 lines
arch/powerpc/kernel/ftrace.c +4 -0
 	}
 
 	return __ftrace_make_call(rec, addr);
+#else
+	/* We should not get here without modules */
+	return -EINVAL;
+#endif /* CONFIG_MODULES */
 }
 
 int ftrace_update_ftrace_func(ftrace_func_t func)
Signature · 3 lines
1.5.6.5

-- 
author maintainer
#8 lore raw

Re: [PATCH 2/2] powerpc, ftrace: fix compile error when modules not configured

Quoted 7 lines · Michael Neuling
Quoted 6 lines · In message <[email protected]> you wrote:

From: Steven Rostedt <[email protected]>

Michael Neuling reported a compile bug when dynamic ftrace was configured in and modules were not. This was due to the ftrace code referencing module specific structures.

Quoted 6 lines
Quoted 2 lines

Steve,

Thanks, fixes the error I was seeing.

Cool.

Quoted 4 lines

As an aside, is there anyway we can merge some of the code in arch/powerpc/kernel/ftrace.c between 32 and 64bit? There seems to be a lot of repeated code in there with only minor changes.

It looks similar, but there's enough differences that I would like to keep them separate. The code is complex enough and has a lot of subtle differences between the two archs.

-- Steve

#9 lore raw
In reply to Steven Rostedt
Quoted 6 lines · Steven Rostedt

Paul,

The first patch is a ftrace clean up to use the proper pr_debug macro.

The second patch is to fix a compile error when dynamic ftrace is compiled in, but modules are not.

Did you find out the problem with the large USB module ?

Cheers,
Ben.

Quoted 14 lines

The following patches are in:

git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-2.6-trace.git

branch: ppc/ftrace

Steven Rostedt (2):
ftrace, powerpc: replace debug macro with proper pr_deug powerpc, ftrace: fix compile error when modules not configured

----

 arch/powerpc/kernel/ftrace.c |   42 +++++++++++++++++++++++++-----------------
 1 files changed, 25 insertions(+), 17 deletions(-)
author maintainer
#10 lore raw
Quoted 9 lines · Benjamin Herrenschmidt
Quoted 6 lines · Steven Rostedt

Paul,

The first patch is a ftrace clean up to use the proper pr_debug macro.

The second patch is to fix a compile error when dynamic ftrace is compiled in, but modules are not.

Did you find out the problem with the large USB module ?

Not yet, I'm still trying to figure out why it does things differently on large modules.

-- Steve

#11 lore raw

hrtimer: about hres_active

In reply to Steven Rostedt

Hi

I am trying to understand the purpose of "hres_active" of hrtimer and have the following question in this regard.

It seems "hres_active" indicates whether high resolution mode is active or not. But I am not clear about the idea behind it.

I see that hres_active is initialized to 0 here:
hrtimer_init_hres

and set to 1 here:
hrtimer_run_pending

      -> hrtimer_switch_to_hres

That means hrtimer becomes "active" at the 1st timer softirq and remains so forever. Is this understanding correct?

My original concern is as follows:

hrtimer_get_next_event returns KTIME_MAX when hrtimer is "active". So if the above understanding is correct, then after the 1st timer softirq it will always return KTIME_MAX. This means cpu_idle will never take the hrtimer event into account and will always base its decision on the next event of the timer wheel. Is this intended behaviour?

I would highly appreciate any information about this. Please CC me because I am not a member of this ML.

Best regards
Iram

#12 lore raw

Re: hrtimer: about hres_active

Iram,

Quoted 9 lines · Iram Shahzad

I am trying to understand the purpose of "hres_active" of hrtimer and have the following question in this regard.

It seems "hres_active" indicates whether high resolution mode is active or not. But I am not clear about the idea behind it.

I see that hres_active is initialized to 0 here:
hrtimer_init_hres

Quoted 6 lines

and set to 1 here:

   hrtimer_run_pending
      -> hrtimer_switch_to_hres

That means hrtimer becomes "active" at the 1st timer softirq and remains so forever. Is this understanding correct?

No. The system switches to high resolution mode late in the boot process and it does so only when there is high res capable hardware available.

Quoted 7 lines

My original concern is as follows:

hrtimer_get_next_event returns KTIME_MAX when hrtimer is "active". So if the above understanding is correct, then after the 1st timer softirq it will always return KTIME_MAX. This means cpu_idle will never take the hrtimer event into account and will always base its decision on the next event of the timer wheel. Is this intended behaviour?

Yes it is. In the case of high res active the hrtimer which is the next to expire is already armed on that CPU in the clock event device. Therefor we know already when the next hrtimer will fire. No need to lookup further. But we have to check the timer wheel as it might have a timer which is due earlier than the first to expire hrtimer.

Thanks,

tglx

#13 lore raw

Re: hrtimer: about hres_active

Thomas,

Many thanks for the reply. I understood why hrtimer list is not checked in hrtimer_get_next_event in the case of high res active.

Please let me ask more about the following.

Quoted 3 lines

No. The system switches to high resolution mode late in the boot process and it does so only when there is high res capable hardware available.

So far as I checked, hres_active is set to 1 in hrtimer_switch_to_hres. hrtimer_switch_to_hres is only called from hrtimer_run_pending. And hrtimer_run_pending is called from timer wheel softirq. That is why I concluded that hres_active is set at 1st timer softirq. Do you just mean that it is not the 1st softirq, or do you mean that in some other place in the late boot process
there is another code which sets hres_active?

And is it correct that hres_active remains 1 forever once set to it?

Best regards
Iram

----- Original Message ----- From: "Thomas Gleixner" <[email protected]>
To: "Iram Shahzad" <[email protected]>
Cc: <[email protected]>
Sent: Tuesday, May 11, 2010 12:12 AM
Subject: Re: hrtimer: about hres_active

Quoted 43 lines

Iram,

Quoted 9 lines · Iram Shahzad

I am trying to understand the purpose of "hres_active" of hrtimer and have the following question in this regard.

It seems "hres_active" indicates whether high resolution mode is active or not. But I am not clear about the idea behind it.

I see that hres_active is initialized to 0 here:
hrtimer_init_hres

Quoted 6 lines

and set to 1 here:

   hrtimer_run_pending
      -> hrtimer_switch_to_hres

That means hrtimer becomes "active" at the 1st timer softirq and remains so forever. Is this understanding correct?

No. The system switches to high resolution mode late in the boot process and it does so only when there is high res capable hardware available.

Quoted 7 lines

My original concern is as follows:

hrtimer_get_next_event returns KTIME_MAX when hrtimer is "active". So if the above understanding is correct, then after the 1st timer softirq it will always return KTIME_MAX. This means cpu_idle will never take the hrtimer event into account and will always base its decision on the next event of the timer wheel. Is this intended behaviour?

Yes it is. In the case of high res active the hrtimer which is the next to expire is already armed on that CPU in the clock event device. Therefor we know already when the next hrtimer will fire. No need to lookup further. But we have to check the timer wheel as it might have a timer which is due earlier than the first to expire hrtimer.

Thanks,

tglx

#14 lore raw

Re: hrtimer: about hres_active

Iram,

please do _NOT_ top post.

Quoted 11 lines · Iram Shahzad
Quoted 3 lines

No. The system switches to high resolution mode late in the boot process and it does so only when there is high res capable hardware available.

So far as I checked, hres_active is set to 1 in hrtimer_switch_to_hres. hrtimer_switch_to_hres is only called from hrtimer_run_pending. And hrtimer_run_pending is called from timer wheel softirq. That is why I concluded that hres_active is set at 1st timer softirq. Do you just mean that it is not the 1st softirq, or do you mean that in some other place in the late boot process
there is another code which sets hres_active?

It happens from softirq context, but not from the 1st softirq.

Quoted 1 line

And is it correct that hres_active remains 1 forever once set to it?

Yes. There is no way back.

Thanks,

tglx