2021-05-29 02:08:11

by Bixuan Cui

[permalink] [raw]
Subject: [PATCH -next] module: fix build error when CONFIG_SMP is disabled

Fix build error when disable CONFIG_SMP:
mm/pgtable-generic.o: In function `.L19':
pgtable-generic.c:(.text+0x42): undefined reference to `flush_pmd_tlb_range'
mm/pgtable-generic.o: In function `pmdp_huge_clear_flush':
pgtable-generic.c:(.text+0x6c): undefined reference to `flush_pmd_tlb_range'
mm/pgtable-generic.o: In function `pmdp_invalidate':
pgtable-generic.c:(.text+0x162): undefined reference to `flush_pmd_tlb_range'

Fixes: e88b333142e4 ("riscv: mm: add THP support on 64-bit")
Reported-by: Hulk Robot <[email protected]>
Signed-off-by: Bixuan Cui <[email protected]>
---
arch/riscv/include/asm/tlbflush.h | 6 ++++++
1 file changed, 6 insertions(+)

diff --git a/arch/riscv/include/asm/tlbflush.h b/arch/riscv/include/asm/tlbflush.h
index c84218ad7afc..9d5801f6e2c3 100644
--- a/arch/riscv/include/asm/tlbflush.h
+++ b/arch/riscv/include/asm/tlbflush.h
@@ -44,6 +44,12 @@ static inline void flush_tlb_range(struct vm_area_struct *vma,
local_flush_tlb_all();
}

+static inline void flush_pmd_tlb_range(struct vm_area_struct *vma,
+ unsigned long start, unsigned long end)
+{
+ local_flush_tlb_all();
+}
+
#define flush_tlb_mm(mm) flush_tlb_all()
#endif /* !CONFIG_SMP || !CONFIG_MMU */

--
2.17.1


2021-05-29 04:46:47

by Nanyong Sun

[permalink] [raw]
Subject: Re: [PATCH -next] module: fix build error when CONFIG_SMP is disabled

On 2021/5/29 10:03, Bixuan Cui wrote:

> diff --git a/arch/riscv/include/asm/tlbflush.h b/arch/riscv/include/asm/tlbflush.h
> index c84218ad7afc..9d5801f6e2c3 100644
> --- a/arch/riscv/include/asm/tlbflush.h
> +++ b/arch/riscv/include/asm/tlbflush.h
> @@ -44,6 +44,12 @@ static inline void flush_tlb_range(struct vm_area_struct *vma,
> local_flush_tlb_all();
> }
>
> +static inline void flush_pmd_tlb_range(struct vm_area_struct *vma,
> + unsigned long start, unsigned long end)
> +{
> + local_flush_tlb_all();
> +}
> +
> #define flush_tlb_mm(mm) flush_tlb_all()
> #endif /* !CONFIG_SMP || !CONFIG_MMU */

Move the prototype of flush_pmd_tlb_range from pgtable.h to tlbflush.h

can also fix this problem, and it seems that declare flush_pmd_tlb_range
in tlbflush.h

is better.

We could

diff --git a/arch/riscv/include/asm/pgtable.h
b/arch/riscv/include/asm/pgtable.h
index 29e2c836848d..eda31002c4b1 100644
--- a/arch/riscv/include/asm/pgtable.h
+++ b/arch/riscv/include/asm/pgtable.h
@@ -622,11 +622,6 @@ static inline pmd_t pmdp_establish(struct
vm_area_struct *vma,
 {
        return __pmd(atomic_long_xchg((atomic_long_t *)pmdp,
pmd_val(pmd)));
 }
-
-#define __HAVE_ARCH_FLUSH_PMD_TLB_RANGE
-void flush_pmd_tlb_range(struct vm_area_struct *vma, unsigned long start,
-                       unsigned long end);
-
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */

 /*
diff --git a/arch/riscv/include/asm/tlbflush.h
b/arch/riscv/include/asm/tlbflush.h
index c84218ad7afc..b9b9c64ac974 100644
--- a/arch/riscv/include/asm/tlbflush.h
+++ b/arch/riscv/include/asm/tlbflush.h
@@ -33,6 +33,11 @@ void flush_tlb_mm(struct mm_struct *mm);
 void flush_tlb_page(struct vm_area_struct *vma, unsigned long addr);
 void flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
                     unsigned long end);
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+#define __HAVE_ARCH_FLUSH_PMD_TLB_RANGE
+void flush_pmd_tlb_range(struct vm_area_struct *vma, unsigned long start,
+                       unsigned long end);
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
 #else /* CONFIG_SMP && CONFIG_MMU */

 #define flush_tlb_all() local_flush_tlb_all()

In this way , flush_p(m/u)d_tlb_range will be defined in
include/linux/pgtable.h if

CONFIG_SMP is not set.

2021-05-29 06:22:44

by Bixuan Cui

[permalink] [raw]
Subject: Re: [PATCH -next] module: fix build error when CONFIG_SMP is disabled



On 2021/5/29 12:43, Sunnanyong (Nanyong Sun, Intelligent Computing Solution Development Dep) wrote:
> Move the prototype of flush_pmd_tlb_range from pgtable.h to tlbflush.h
>
> can also fix this problem, and it seems that declare flush_pmd_tlb_range in tlbflush.h
>
> is better.
>
> We could
Agree with you.

Thanks
Bixuan Cui