Re: [PATCHv4] mm: Don't offset memmap for flatmem

[Vlastimil Babka <vbabka@suse.cz>]

On 10/07/2015 11:28 PM, Laura Abbott wrote:
> From: Laura Abbott <laura@labbott.name>
> 
> Srinivas Kandagatla reported bad page messages when trying to
> remove the bottom 2MB on an ARM based IFC6410 board
> 
> BUG: Bad page state in process swapper  pfn:fffa8
> page:ef7fb500 count:0 mapcount:0 mapping:  (null) index:0x0
> flags: 0x96640253(locked|error|dirty|active|arch_1|reclaim|mlocked)
> page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set
> bad because of flags:
> flags: 0x200041(locked|active|mlocked)
> Modules linked in:
> CPU: 0 PID: 0 Comm: swapper Not tainted 3.19.0-rc3-00007-g412f9ba-dirty #816
> Hardware name: Qualcomm (Flattened Device Tree)
> [<c0218280>] (unwind_backtrace) from [<c0212be8>] (show_stack+0x20/0x24)
> [<c0212be8>] (show_stack) from [<c0af7124>] (dump_stack+0x80/0x9c)
> [<c0af7124>] (dump_stack) from [<c0301570>] (bad_page+0xc8/0x128)
> [<c0301570>] (bad_page) from [<c03018a8>] (free_pages_prepare+0x168/0x1e0)
> [<c03018a8>] (free_pages_prepare) from [<c030369c>] (free_hot_cold_page+0x3c/0x174)
> [<c030369c>] (free_hot_cold_page) from [<c0303828>] (__free_pages+0x54/0x58)
> [<c0303828>] (__free_pages) from [<c030395c>] (free_highmem_page+0x38/0x88)
> [<c030395c>] (free_highmem_page) from [<c0f62d5c>] (mem_init+0x240/0x430)
> [<c0f62d5c>] (mem_init) from [<c0f5db3c>] (start_kernel+0x1e4/0x3c8)
> [<c0f5db3c>] (start_kernel) from [<80208074>] (0x80208074)
> Disabling lock debugging due to kernel taint
> 
> Removing the lower 2MB made the start of the lowmem zone to no longer
> be page block aligned. IFC6410 uses CONFIG_FLATMEM where
> alloc_node_mem_map allocates memory for the mem_map. alloc_node_mem_map
> will offset for unaligned nodes with the assumption the pfn/page
> translation functions will account for the offset. The functions for
> CONFIG_FLATMEM do not offset however, resulting in overrunning
> the memmap array. Just use the allocated memmap without any offset
> when running with CONFIG_FLATMEM to avoid the overrun.
> 
> Signed-off-by: Laura Abbott <laura@labbott.name>
> Signed-off-by: Laura Abbott <lauraa@codeaurora.org>
> Reported-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
> Tested-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
> Acked-by: Vlastimil Babka <vbabka@suse.cz>
> Tested-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
> 
> 
> ---
> I was reminded at Linaro Connect that I never followed up on this patch.
> v3 (http://marc.info/?l=linux-mm&m=142629173021316&w=2)
> was dropped due to crashing (https://lkml.org/lkml/2015/3/20/546)
> The fundamentals of the patch were fine, the VM_BUG_ON I added turned out
> to be incorrect. It was supposed to be a sanity check but it didn't
> actually hold true in all cases (or even most cases...)

Yeah, sorry about the VM_BUG_ON (I think I suggested it originally). It's indeed
wrong, I can see it now... the self-referentiality of "page_to_pfn(mem_map)" on
FLATMEM makes my head hurt.

> v4 just drops
> the sanity check. I verified that the offset is working on both ARM, x86,
> and i386 QEMU with the config provided by Steven Rostedt.

Great, thanks.

> Since this was minor diff from the previous patch, I kept all the previous
> Acked-by/Tested-by.
> ---
>  mm/page_alloc.c | 9 ++++++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 40e2942..9e5ba5e 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4953,6 +4953,8 @@ static void __paginginit free_area_init_core(struct pglist_data *pgdat,
>  
>  static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat)
>  {
> +	unsigned long __maybe_unused offset = 0;
> +
>  	/* Skip empty nodes */
>  	if (!pgdat->node_spanned_pages)
>  		return;
> @@ -4969,6 +4971,7 @@ static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat)
>  		 * for the buddy allocator to function correctly.
>  		 */
>  		start = pgdat->node_start_pfn & ~(MAX_ORDER_NR_PAGES - 1);
> +		offset = pgdat->node_start_pfn - start;
>  		end = pgdat_end_pfn(pgdat);
>  		end = ALIGN(end, MAX_ORDER_NR_PAGES);
>  		size =  (end - start) * sizeof(struct page);
> @@ -4976,7 +4979,7 @@ static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat)
>  		if (!map)
>  			map = memblock_virt_alloc_node_nopanic(size,
>  							       pgdat->node_id);
> -		pgdat->node_mem_map = map + (pgdat->node_start_pfn - start);
> +		pgdat->node_mem_map = map + offset;
>  	}
>  #ifndef CONFIG_NEED_MULTIPLE_NODES
>  	/*
> @@ -4984,9 +4987,9 @@ static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat)
>  	 */
>  	if (pgdat == NODE_DATA(0)) {
>  		mem_map = NODE_DATA(0)->node_mem_map;
> -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP
> +#if defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP) || defined(CONFIG_FLATMEM)
>  		if (page_to_pfn(mem_map) != pgdat->node_start_pfn)
> -			mem_map -= (pgdat->node_start_pfn - ARCH_PFN_OFFSET);
> +			mem_map -= offset;
>  #endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */
>  	}
>  #endif
>