Skip to contentSkip to conversation

parisc: move CPU field back into thread_info

PATCH In review Ard Biesheuvel 3 participants 6 messages Nov 4, 2021 PA-RISC architecture

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 parisc: move CPU field back into thread_info

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")
---
 arch/parisc/include/asm/smp.h         | 19 ++-----------------
 arch/parisc/include/asm/thread_info.h |  2 ++
 arch/parisc/kernel/asm-offsets.c      |  4 +---
 arch/parisc/kernel/smp.c              |  2 +-
 4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
 #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()		(0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()		(*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
 #else /* CONFIG_SMP */
 
 static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
 struct thread_info {
 	unsigned long flags;		/* thread_info flags (see TIF_*) */
+	__u32 cpu;			/* current CPU */
 	int preempt_count;		/* 0=premptable, <0=BUG; will also serve as bh-counter */
 };
 
 #define INIT_THREAD_INFO(tsk)			\
 {						\
 	.flags		= 0,			\
+	.cpu		= 0,			\
 	.preempt_count	= INIT_PREEMPT_COUNT,	\
 }
 
arch/parisc/kernel/asm-offsets.c +1 -3
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
  *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
  */
 
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/thread_info.h>
@@ -40,7 +38,7 @@ int main(void)
 	DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
 	DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
 #ifdef CONFIG_SMP
-	DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+	DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
 #endif
 	BLANK();
 	DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +1 -1
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
 	const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
 	long timeout;
 
-	idle->cpu = cpuid;
+	task_thread_info(idle)->cpu = cpuid;
 
 	/* Let _start know what logical CPU we're booting
 	** (offset into init_tasks[],cpu_data[])
Signature · 1 line
2.30.2
maintainer
#1 lore raw
Quoted 6 lines · Ard Biesheuvel

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

Quoted 7 lines

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")

Thanks Ard!

I actually had the same patch finished right now too :-)

One small nid though. See below...

Quoted 70 lines
---
 arch/parisc/include/asm/smp.h         | 19 ++-----------------
 arch/parisc/include/asm/thread_info.h |  2 ++
 arch/parisc/kernel/asm-offsets.c      |  4 +---
 arch/parisc/kernel/smp.c              |  2 +-
 4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
 #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()		(0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()		(*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
 #else /* CONFIG_SMP */
 
 static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
 struct thread_info {
 	unsigned long flags;		/* thread_info flags (see TIF_*) */
+	__u32 cpu;			/* current CPU */
 	int preempt_count;		/* 0=premptable, <0=BUG; will also serve as bh-counter */
 };
 
 #define INIT_THREAD_INFO(tsk)			\
 {						\
 	.flags		= 0,			\
+	.cpu		= 0,			\
 	.preempt_count	= INIT_PREEMPT_COUNT,	\
 }
 
arch/parisc/kernel/asm-offsets.c +0 -2
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
  *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
  */
 
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/thread_info.h>

arch/parisc/kernel/asm-offsets.c int main(void) · line 42

Quoted 6 lines
arch/parisc/kernel/asm-offsets.c +1 -1
@@ -40,7 +38,7 @@ int main(void)
 	DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
 	DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
 #ifdef CONFIG_SMP
-	DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+	DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
 #endif

The TASK_CPU define isn't needed any longer.
So, the whole part inside #ifdef..#endif can go.

If it's Ok for you I'll clean up your patch and submit all with the next pull request to Linus later today.

Helge

Quoted 16 lines
arch/parisc/kernel/asm-offsets.c +0 -0
 	BLANK();
 	DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +1 -1
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
 	const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
 	long timeout;
 
-	idle->cpu = cpuid;
+	task_thread_info(idle)->cpu = cpuid;
 
 	/* Let _start know what logical CPU we're booting
 	** (offset into init_tasks[],cpu_data[])
#2 lore raw
Quoted 12 lines · Helge Deller
Quoted 10 lines · Ard Biesheuvel

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Quoted 8 lines
Quoted 3 lines
Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")

Thanks Ard!

I actually had the same patch finished right now too :-)

I was wondering about that :-)

Quoted 67 lines

One small nid though. See below...

Quoted 64 lines
---
 arch/parisc/include/asm/smp.h         | 19 ++-----------------
 arch/parisc/include/asm/thread_info.h |  2 ++
 arch/parisc/kernel/asm-offsets.c      |  4 +---
 arch/parisc/kernel/smp.c              |  2 +-
 4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
 #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()               (0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()               (*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
 #else /* CONFIG_SMP */
 
 static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
 struct thread_info {
      unsigned long flags;            /* thread_info flags (see TIF_*) */
+     __u32 cpu;                      /* current CPU */
      int preempt_count;              /* 0=premptable, <0=BUG; will also serve as bh-counter */
 };
 
 #define INIT_THREAD_INFO(tsk)                        \
 {                                            \
      .flags          = 0,                    \
+     .cpu            = 0,                    \
      .preempt_count  = INIT_PREEMPT_COUNT,   \
 }
 
arch/parisc/kernel/asm-offsets.c +0 -0
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
  *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
  */
 
Quoted 16 lines

arch/parisc/kernel/asm-offsets.c int main(void) · line 42

Quoted 12 lines
arch/parisc/kernel/asm-offsets.c +1 -3
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/thread_info.h>
@@ -40,7 +38,7 @@ int main(void)
      DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
      DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
 #ifdef CONFIG_SMP
-     DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+     DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
 #endif

The TASK_CPU define isn't needed any longer.
So, the whole part inside #ifdef..#endif can go.

OK let's just drop it then.

Quoted 10 lines
Quoted 10 lines
arch/parisc/kernel/asm-offsets.c +0 -0
      BLANK();
      DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +0 -0
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
      const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
      long timeout;
 
Quoted 2 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle) · line 342

Quoted 2 lines
arch/parisc/kernel/smp.c +1 -1
-     idle->cpu = cpuid;
+     task_thread_info(idle)->cpu = cpuid;

I was wondering if this could simply be dropped altogether? I am pretty sure it is redundant, as the core code sets this field as well, but I don't have access to a SMP PA-RISC machine to test it.

Quoted 8 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle)

Quoted 4 lines
arch/parisc/kernel/smp.c +0 -0
      /* Let _start know what logical CPU we're booting
      ** (offset into init_tasks[],cpu_data[])
 

If it's Ok for you I'll clean up your patch and submit all with the next pull request to Linus later today.

Whatever works for you is fine with me.

maintainer
#3 lore raw
Quoted 108 lines · Ard Biesheuvel
Quoted 12 lines · Helge Deller
Quoted 10 lines · Ard Biesheuvel

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Quoted 8 lines
Quoted 3 lines
Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")

Thanks Ard!

I actually had the same patch finished right now too :-)

I was wondering about that :-)

Quoted 67 lines

One small nid though. See below...

Quoted 64 lines
---
 arch/parisc/include/asm/smp.h         | 19 ++-----------------
 arch/parisc/include/asm/thread_info.h |  2 ++
 arch/parisc/kernel/asm-offsets.c      |  4 +---
 arch/parisc/kernel/smp.c              |  2 +-
 4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
 #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()               (0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()               (*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
 #else /* CONFIG_SMP */
 
 static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
 struct thread_info {
      unsigned long flags;            /* thread_info flags (see TIF_*) */
+     __u32 cpu;                      /* current CPU */
      int preempt_count;              /* 0=premptable, <0=BUG; will also serve as bh-counter */
 };
 
 #define INIT_THREAD_INFO(tsk)                        \
 {                                            \
      .flags          = 0,                    \
+     .cpu            = 0,                    \
      .preempt_count  = INIT_PREEMPT_COUNT,   \
 }
 
arch/parisc/kernel/asm-offsets.c +0 -0
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
  *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
  */
 
Quoted 16 lines

arch/parisc/kernel/asm-offsets.c int main(void) · line 42

Quoted 12 lines
arch/parisc/kernel/asm-offsets.c +1 -3
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/thread_info.h>
@@ -40,7 +38,7 @@ int main(void)
      DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
      DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
 #ifdef CONFIG_SMP
-     DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+     DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
 #endif

The TASK_CPU define isn't needed any longer.
So, the whole part inside #ifdef..#endif can go.

Quoted 1 line

OK let's just drop it then.

Ok

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
arch/parisc/kernel/asm-offsets.c +0 -0
      BLANK();
      DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +0 -0
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
      const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
      long timeout;
 
Quoted 6 lines
Quoted 2 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle) · line 342

Quoted 2 lines
arch/parisc/kernel/smp.c +1 -1
-     idle->cpu = cpuid;
+     task_thread_info(idle)->cpu = cpuid;

I was wondering if this could simply be dropped altogether? I am pretty sure it is redundant, as the core code sets this field as well, but I don't have access to a SMP PA-RISC machine to test it.

Good point.
I tested and it can be dropped.

Quoted 9 lines
Quoted 7 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle)

Quoted 3 lines
arch/parisc/kernel/smp.c +0 -0
      /* Let _start know what logical CPU we're booting
      ** (offset into init_tasks[],cpu_data[])
 

If it's Ok for you I'll clean up your patch and submit all with the next pull request to Linus later today.

Whatever works for you is fine with me.

Ok. I've cleaned it up, applied it to my for-next tree and will push later today to Linus with other patches. Here is the current version:
https://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux.git/commit/?h=for-next

Thanks!
Helge

#4 lore raw
Quoted 145 lines · Helge Deller
Quoted 108 lines · Ard Biesheuvel
Quoted 12 lines · Helge Deller
Quoted 10 lines · Ard Biesheuvel

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Quoted 8 lines
Quoted 3 lines
Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")

Thanks Ard!

I actually had the same patch finished right now too :-)

I was wondering about that :-)

Quoted 67 lines

One small nid though. See below...

Quoted 64 lines
---
 arch/parisc/include/asm/smp.h         | 19 ++-----------------
 arch/parisc/include/asm/thread_info.h |  2 ++
 arch/parisc/kernel/asm-offsets.c      |  4 +---
 arch/parisc/kernel/smp.c              |  2 +-
 4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
 #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()               (0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()               (*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
 #else /* CONFIG_SMP */
 
 static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
 struct thread_info {
      unsigned long flags;            /* thread_info flags (see TIF_*) */
+     __u32 cpu;                      /* current CPU */
      int preempt_count;              /* 0=premptable, <0=BUG; will also serve as bh-counter */
 };
 
 #define INIT_THREAD_INFO(tsk)                        \
 {                                            \
      .flags          = 0,                    \
+     .cpu            = 0,                    \
      .preempt_count  = INIT_PREEMPT_COUNT,   \
 }
 
arch/parisc/kernel/asm-offsets.c +0 -0
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
  *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
  */
 
Quoted 16 lines

arch/parisc/kernel/asm-offsets.c int main(void) · line 42

Quoted 12 lines
arch/parisc/kernel/asm-offsets.c +1 -3
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/thread_info.h>
@@ -40,7 +38,7 @@ int main(void)
      DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
      DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
 #ifdef CONFIG_SMP
-     DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+     DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
 #endif

The TASK_CPU define isn't needed any longer.
So, the whole part inside #ifdef..#endif can go.

Quoted 1 line

OK let's just drop it then.

Ok

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
arch/parisc/kernel/asm-offsets.c +0 -0
      BLANK();
      DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +0 -0
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
      const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
      long timeout;
 
Quoted 6 lines
Quoted 2 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle) · line 342

Quoted 2 lines
arch/parisc/kernel/smp.c +1 -1
-     idle->cpu = cpuid;
+     task_thread_info(idle)->cpu = cpuid;

I was wondering if this could simply be dropped altogether? I am pretty sure it is redundant, as the core code sets this field as well, but I don't have access to a SMP PA-RISC machine to test it.

Good point.
I tested and it can be dropped.

Quoted 8 lines
Quoted 7 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle)

Quoted 3 lines
arch/parisc/kernel/smp.c +0 -0
      /* Let _start know what logical CPU we're booting
      ** (offset into init_tasks[],cpu_data[])
 

If it's Ok for you I'll clean up your patch and submit all with the next pull request to Linus later today.

Quoted 6 lines
Quoted 1 line

Whatever works for you is fine with me.

Ok. I've cleaned it up, applied it to my for-next tree and will push later today to Linus with other patches. Here is the current version:
https://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux.git/commit/?h=for-next

Looks fine

Thanks,

#5 lore raw
In reply to Helge Deller
Quoted 144 lines · Helge Deller
Quoted 108 lines · Ard Biesheuvel
Quoted 12 lines · Helge Deller
Quoted 10 lines · Ard Biesheuvel

In commit 2214c0e77259 ("parisc: Move thread_info into task struct") PA-RISC gained support for THREAD_INFO_IN_TASK while changes were already underway to keep the CPU field in thread_info rather than move it into task_struct when THREAD_INFO_IN_TASK is enabled. The result is a broken build for all PA-RISC configs that enable SMP.

So let's partially revert that commit, and get rid of the ugly hack to get at the offset of task_struct::cpu without having to include linux/sched.h, and put the CPU field back where it was before.

Quoted 8 lines
Quoted 3 lines
Fixes bcf9033e5449 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")

Thanks Ard!

I actually had the same patch finished right now too :-)

I was wondering about that :-)

Quoted 67 lines

One small nid though. See below...

Quoted 64 lines
---
  arch/parisc/include/asm/smp.h         | 19 ++-----------------
  arch/parisc/include/asm/thread_info.h |  2 ++
  arch/parisc/kernel/asm-offsets.c      |  4 +---
  arch/parisc/kernel/smp.c              |  2 +-
  4 files changed, 6 insertions(+), 21 deletions(-)
arch/parisc/include/asm/smp.h +2 -17
index 16d41127500e..cba55abf7bac 100644
@@ -34,23 +34,8 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
 
  #endif /* !ASSEMBLY */
 
-/*
- * This is particularly ugly: it appears we can't actually get the definition
- * of task_struct here, but we need access to the CPU this task is running on.
- * Instead of using task_struct we're using TASK_CPU which is extracted from
- * asm-offsets.h by kbuild to get the current processor ID.
- *
- * This also needs to be safeguarded when building asm-offsets.s because at
- * that time TASK_CPU is not defined yet. It could have been guarded by
- * TASK_CPU itself, but we want the build to fail if TASK_CPU is missing
- * when building something else than asm-offsets.s
- */
-#ifdef GENERATING_ASM_OFFSETS
-#define raw_smp_processor_id()               (0)
-#else
-#include <asm/asm-offsets.h>
-#define raw_smp_processor_id()               (*(unsigned int *)((void *)current + TASK_CPU))
-#endif
+#define raw_smp_processor_id() (current_thread_info()->cpu)
+
  #else /* CONFIG_SMP */
 
  static inline void smp_send_all_nop(void) { return; }
arch/parisc/include/asm/thread_info.h +2 -0
index 75657c2c54e1..d6268834bfa5 100644
@@ -8,12 +8,14 @@
 
  struct thread_info {
       unsigned long flags;            /* thread_info flags (see TIF_*) */
+     __u32 cpu;                      /* current CPU */
       int preempt_count;              /* 0=premptable, <0=BUG; will also serve as bh-counter */
  };
 
  #define INIT_THREAD_INFO(tsk)                        \
  {                                            \
       .flags          = 0,                    \
+     .cpu            = 0,                    \
       .preempt_count  = INIT_PREEMPT_COUNT,   \
  }
 
arch/parisc/kernel/asm-offsets.c +0 -0
index e35154035441..629d38e36e22 100644
@@ -14,8 +14,6 @@
   *    Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org>
   */
 
Quoted 16 lines

arch/parisc/kernel/asm-offsets.c int main(void) · line 42

Quoted 12 lines
arch/parisc/kernel/asm-offsets.c +1 -3
-#define GENERATING_ASM_OFFSETS /* asm/smp.h */
-
  #include <linux/types.h>
  #include <linux/sched.h>
  #include <linux/thread_info.h>
@@ -40,7 +38,7 @@ int main(void)
       DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags));
       DEFINE(TASK_STACK, offsetof(struct task_struct, stack));
  #ifdef CONFIG_SMP
-     DEFINE(TASK_CPU, offsetof(struct task_struct, cpu));
+     DEFINE(TASK_CPU, offsetof(struct task_struct, thread_info.cpu));
  #endif

The TASK_CPU define isn't needed any longer.
So, the whole part inside #ifdef..#endif can go.

Quoted 1 line

OK let's just drop it then.

Ok

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
arch/parisc/kernel/asm-offsets.c +0 -0
       BLANK();
       DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs));
arch/parisc/kernel/smp.c +0 -0
index 171925285f3e..0cd97fa004c5 100644
@@ -339,7 +339,7 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle)
       const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid);
       long timeout;
 
Quoted 6 lines
Quoted 2 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle) · line 342

Quoted 2 lines
arch/parisc/kernel/smp.c +1 -1
-     idle->cpu = cpuid;
+     task_thread_info(idle)->cpu = cpuid;

I was wondering if this could simply be dropped altogether? I am pretty sure it is redundant, as the core code sets this field as well, but I don't have access to a SMP PA-RISC machine to test it.

Good point.
I tested and it can be dropped.

Quoted 8 lines
Quoted 7 lines

arch/parisc/kernel/smp.c int smp_boot_one_cpu(int cpuid, struct task_struct *idle)

Quoted 3 lines
arch/parisc/kernel/smp.c +0 -0
       /* Let _start know what logical CPU we're booting
       ** (offset into init_tasks[],cpu_data[])
 

If it's Ok for you I'll clean up your patch and submit all with the next pull request to Linus later today.

Quoted 6 lines
Quoted 1 line

Whatever works for you is fine with me.

Ok. I've cleaned it up, applied it to my for-next tree and will push later today to Linus with other patches. Here is the current version:
https://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux.git/commit/?h=for-next

Works for me.

Thanks,
Guenter