[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20150804121801.GA28571@dhcp22.suse.cz>
Date: Tue, 4 Aug 2015 14:18:01 +0200
From: Michal Hocko <mhocko@...nel.org>
To: Guenter Roeck <linux@...ck-us.net>
Cc: Vladimir Davydov <vdavydov@...allels.com>,
Andrew Morton <akpm@...ux-foundation.org>,
linux-kernel@...r.kernel.org,
"David S. Miller" <davem@...emloft.net>, sparclinux@...r.kernel.org
Subject: Re: Build failure (sparc32:allmodconfig) in -next due to 'memcg:
export struct mem_cgroup'
On Sat 01-08-15 20:02:20, Guenter Roeck wrote:
> Hi,
>
> Commit b65390c6c2178f ("memcg: export struct mem_cgroup") causes several
> build failures in -next (next-20150731) when trying to build
> sparc32:allmodconfig.
>
> Below are some of the errors, as well as a bisect log.
>
> Guenter
>
> ---
> In file included from include/linux/pid_namespace.h:6:0,
> from include/linux/ptrace.h:8,
> from include/linux/ftrace.h:13,
> from include/linux/init_task.h:8,
> from init/init_task.c:1:
> include/linux/mm.h: At top level:
> include/linux/mm.h:888:7: error: conflicting types for 'page_address'
> ./arch/sparc/include/asm/highmem.h:59:10: note: previous implicit declaration of 'page_address' was here
This is weird:
* include/linux/mm.h:888
#if defined(HASHED_PAGE_VIRTUAL)
void *page_address(const struct page *page);
void set_page_address(struct page *page, void *virtual);
void page_address_init(void);
#endif
* arch/sparc/include/asm/highmem.h:59
static inline void *kmap(struct page *page)
{
BUG_ON(in_interrupt());
if (!PageHighMem(page))
return page_address(page);
return kmap_high(page);
}
How come we have conflicting types here. page_address is clearly not
a function declaration in sparc code. Maybe something expands to an
incomplete c construct.
I have tried to crosscompile with
https://www.kernel.org/pub/tools/crosstool/files/bin/x86_64/4.6.3/ but
failed with strange errors.
$ PATH=/mnt/share/devel/miso/crosstool/sparc/gcc-4.6.3-nolibc/x86_64-linux/bin:$PATH CROSS_COMPILE=x86_64-linux- ARCH=sparc32 make allmodconfig
warning: (USB_OTG_FSM && FSL_USB2_OTG && USB_MV_OTG) selects USB_OTG which has unmet direct dependencies (USB_SUPPORT && USB && PM)
warning: (MFD_CROS_EC) selects CHROME_PLATFORMS which has unmet direct dependencies (X86 || ARM)
warning: (SND_SOC_ROCKCHIP_MAX98090 && SND_SOC_ROCKCHIP_RT5645) selects SND_SOC_ROCKCHIP_I2S which has unmet direct dependencies (SOUND && !M68K && !UML && SND && SND_SOC && CLKDEV_LOOKUP && SND_SOC_ROCKCHIP)
$ PATH=/mnt/share/devel/miso/crosstool/sparc/gcc-4.6.3-nolibc/x86_64-linux/bin:$PATH CROSS_COMPILE=x86_64-linux- ARCH=sparc32 make
x86_64-linux-gcc: warning: '-mcpu=' is deprecated; use '-mtune=' or '-march=' instead
cc1: error: unrecognized command line option '-mno-fpu'
cc1: warning: unknown register name: g5 [enabled by default]
cc1: warning: unknown register name: g7 [enabled by default]
kernel/bounds.c:1:0: error: bad value (v8) for -mtune= switch
Kbuild:44: recipe for target 'kernel/bounds.s' failed
make[1]: *** [kernel/bounds.s] Error 1
Makefile:1009: recipe for target 'prepare0' failed
make: *** [prepare0] Error 2
So either I am doing something wrong or the cross compiler is broken.
Could you send the preprocessed output of the failing compile?
> In file included from include/linux/highmem.h:7:0,
> from include/linux/bio.h:23,
> from include/linux/writeback.h:192,
> from include/linux/memcontrol.h:30,
> from include/linux/swap.h:8,
> from ./arch/sparc/include/asm/pgtable_32.h:17,
> from ./arch/sparc/include/asm/pgtable.h:6,
> from arch/sparc/kernel/traps_32.c:23:
> include/linux/mm.h: In function 'is_vmalloc_addr':
> include/linux/mm.h:371:17: error: 'VMALLOC_START' undeclared (first use in this function)
> include/linux/mm.h:371:17: note: each undeclared identifier is reported only once for each function it appears in
> include/linux/mm.h:371:41: error: 'VMALLOC_END' undeclared (first use in this function)
> include/linux/mm.h: In function 'maybe_mkwrite':
> include/linux/mm.h:556:3: error: implicit declaration of function 'pte_mkwrite'
Hmm, this is a clear dependency between headers. I am not
sure why this header is even needed. swap_entry is defined in
linux/mm_types.h. Does the code compile if you replace this include by
include <linux/mm_types.h>?
But maybe we should rather get rid of the include and make __swp_type,
__swp_offset and __swp_entry macros and mimic 64b header. Something like
the (untested) patch below:
> In file included from include/linux/highmem.h:7:0,
> from include/linux/bio.h:23,
> from include/linux/writeback.h:192,
> from include/linux/memcontrol.h:30,
> from include/linux/swap.h:8,
> from ./arch/sparc/include/asm/pgtable_32.h:17,
> from ./arch/sparc/include/asm/pgtable.h:6,
> from ./arch/sparc/include/asm/pgalloc_32.h:8,
> from ./arch/sparc/include/asm/pgalloc.h:6,
> from arch/sparc/kernel/sun4m_irq.c:16:
> include/linux/mm.h: In function 'pte_lockptr':
> include/linux/mm.h:1422:2: error: implicit declaration of function 'pmd_page' [-Werror=implicit-function-declaration]
> include/linux/mm.h:1422:2: error: passing argument 1 of 'ptlock_ptr' makes pointer from integer without a cast [-Werror]
> include/linux/mm.h:1396:27: note: expected 'struct page *' but argument is of type 'int'
> include/linux/mm.h: In function 'pgtable_init':
> include/linux/mm.h:1465:2: error: implicit declaration of function 'pgtable_cache_init'
This falls into the same category.
---
diff --git a/arch/sparc/include/asm/pgtable_32.h b/arch/sparc/include/asm/pgtable_32.h
index f06b36a00a3b..df5c9b16ea12 100644
--- a/arch/sparc/include/asm/pgtable_32.h
+++ b/arch/sparc/include/asm/pgtable_32.h
@@ -348,22 +348,14 @@ void srmmu_mapiorange(unsigned int bus, unsigned long xpa,
void srmmu_unmapiorange(unsigned long virt_addr, unsigned int len);
/* Encode and de-code a swap entry */
-static inline unsigned long __swp_type(swp_entry_t entry)
-{
- return (entry.val >> SRMMU_SWP_TYPE_SHIFT) & SRMMU_SWP_TYPE_MASK;
-}
-
-static inline unsigned long __swp_offset(swp_entry_t entry)
-{
- return (entry.val >> SRMMU_SWP_OFF_SHIFT) & SRMMU_SWP_OFF_MASK;
-}
-
-static inline swp_entry_t __swp_entry(unsigned long type, unsigned long offset)
-{
- return (swp_entry_t) {
- (type & SRMMU_SWP_TYPE_MASK) << SRMMU_SWP_TYPE_SHIFT
- | (offset & SRMMU_SWP_OFF_MASK) << SRMMU_SWP_OFF_SHIFT };
-}
+#define __swp_type(entry) (((entry).val >> SRMMU_SWP_TYPE_MASK) & SRMMU_SWP_TYPE_MASK)
+#define __swp_offset(entry) (((entry).val >> SRMMU_SWP_OFF_SHIFT) & SRMMU_SWP_OFF_MASK)
+#define __swp_entry(type, offset) \
+ ( (swp_entry_t) \
+ { \
+ (((type) & SRMMU_SWP_TYPE_MASK) << SRMMU_SWP_TYPE_SHIFT | \
+ ((offset) & SRMMU_SWP_OFF_MASK) << SRMMU_SWP_OFF_SHIFT) \
+ } )
#define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) })
#define __swp_entry_to_pte(x) ((pte_t) { (x).val })
--
Michal Hocko
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists