From: Helge Deller <deller@gmx.de>
To: Mike Rapoport <rppt@linux.ibm.com>,
"James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Matthew Wilcox <willy@infradead.org>,
linux-parisc@vger.kernel.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] parisc: use memblock_alloc() instead of custom get_memblock()
Date: Thu, 21 Feb 2019 11:00:05 +0100 [thread overview]
Message-ID: <5ce80937-a55a-8e79-2575-27d296078d41@gmx.de> (raw)
In-Reply-To: <20190221090752.GA32004@rapoport-lnx>
On 21.02.19 10:07, Mike Rapoport wrote:
> On Tue, Feb 12, 2019 at 05:16:12PM +0200, Mike Rapoport wrote:
>> The get_memblock() function implements custom bottom-up memblock allocator.
>> Setting 'memblock_bottom_up = true' before any memblock allocation is done
>> allows replacing get_memblock() calls with memblock_alloc().
>> Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Helge Deller <deller@gmx.de>
Tested-by: Helge Deller <deller@gmx.de>
Thanks!
Shall I push the patch upstream with the parisc tree?
Helge
>> ---
>> v2: fix allocation alignment
>>
>> arch/parisc/mm/init.c | 52 +++++++++++++++++++--------------------------------
>> 1 file changed, 19 insertions(+), 33 deletions(-)
>>
>> diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
>> index 059187a..d0b1662 100644
>> --- a/arch/parisc/mm/init.c
>> +++ b/arch/parisc/mm/init.c
>> @@ -79,36 +79,6 @@ static struct resource sysram_resources[MAX_PHYSMEM_RANGES] __read_mostly;
>> physmem_range_t pmem_ranges[MAX_PHYSMEM_RANGES] __read_mostly;
>> int npmem_ranges __read_mostly;
>>
>> -/*
>> - * get_memblock() allocates pages via memblock.
>> - * We can't use memblock_find_in_range(0, KERNEL_INITIAL_SIZE) here since it
>> - * doesn't allocate from bottom to top which is needed because we only created
>> - * the initial mapping up to KERNEL_INITIAL_SIZE in the assembly bootup code.
>> - */
>> -static void * __init get_memblock(unsigned long size)
>> -{
>> - static phys_addr_t search_addr __initdata;
>> - phys_addr_t phys;
>> -
>> - if (!search_addr)
>> - search_addr = PAGE_ALIGN(__pa((unsigned long) &_end));
>> - search_addr = ALIGN(search_addr, size);
>> - while (!memblock_is_region_memory(search_addr, size) ||
>> - memblock_is_region_reserved(search_addr, size)) {
>> - search_addr += size;
>> - }
>> - phys = search_addr;
>> -
>> - if (phys)
>> - memblock_reserve(phys, size);
>> - else
>> - panic("get_memblock() failed.\n");
>> -
>> - memset(__va(phys), 0, size);
>> -
>> - return __va(phys);
>> -}
>> -
>> #ifdef CONFIG_64BIT
>> #define MAX_MEM (~0UL)
>> #else /* !CONFIG_64BIT */
>> @@ -321,6 +291,13 @@ static void __init setup_bootmem(void)
>> max_pfn = start_pfn + npages;
>> }
>>
>> + /*
>> + * We can't use memblock top-down allocations because we only
>> + * created the initial mapping up to KERNEL_INITIAL_SIZE in
>> + * the assembly bootup code.
>> + */
>> + memblock_set_bottom_up(true);
>> +
>> /* IOMMU is always used to access "high mem" on those boxes
>> * that can support enough mem that a PCI device couldn't
>> * directly DMA to any physical addresses.
>> @@ -442,7 +419,10 @@ static void __init map_pages(unsigned long start_vaddr,
>> */
>>
>> if (!pmd) {
>> - pmd = (pmd_t *) get_memblock(PAGE_SIZE << PMD_ORDER);
>> + pmd = memblock_alloc(PAGE_SIZE << PMD_ORDER,
>> + PAGE_SIZE << PMD_ORDER);
>> + if (!pmd)
>> + panic("pmd allocation failed.\n");
>> pmd = (pmd_t *) __pa(pmd);
>> }
>>
>> @@ -461,7 +441,10 @@ static void __init map_pages(unsigned long start_vaddr,
>>
>> pg_table = (pte_t *)pmd_address(*pmd);
>> if (!pg_table) {
>> - pg_table = (pte_t *) get_memblock(PAGE_SIZE);
>> + pg_table = memblock_alloc(PAGE_SIZE,
>> + PAGE_SIZE);
>> + if (!pg_table)
>> + panic("page table allocation failed\n");
>> pg_table = (pte_t *) __pa(pg_table);
>> }
>>
>> @@ -700,7 +683,10 @@ static void __init pagetable_init(void)
>> }
>> #endif
>>
>> - empty_zero_page = get_memblock(PAGE_SIZE);
>> + empty_zero_page = memblock_alloc(PAGE_SIZE, PAGE_SIZE);
>> + if (!empty_zero_page)
>> + panic("zero page allocation failed.\n");
>> +
>> }
>>
>> static void __init gateway_init(void)
>> --
>> 2.7.4
>>
>
next prev parent reply other threads:[~2019-02-21 10:00 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-12 15:16 [PATCH v2] parisc: use memblock_alloc() instead of custom get_memblock() Mike Rapoport
2019-02-21 9:07 ` Mike Rapoport
2019-02-21 10:00 ` Helge Deller [this message]
2019-02-21 11:15 ` Mike Rapoport
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5ce80937-a55a-8e79-2575-27d296078d41@gmx.de \
--to=deller@gmx.de \
--cc=James.Bottomley@HansenPartnership.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-parisc@vger.kernel.org \
--cc=rppt@linux.ibm.com \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).