linux-kernel - Re: [RFT/PATCH] x86: Unify kernel_physical_mapping

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Fri, 26 Feb 2010 15:37:24 +0200
From:	Pekka Enberg <penberg@...helsinki.fi>
To:	KAMEZAWA Hiroyuki <kamezawa.hiroyu@...fujitsu.com>
CC:	mingo@...e.hu, x86@...nel.org, linux-kernel@...r.kernel.org,
	yinghai@...nel.org
Subject: Re: [RFT/PATCH] x86: Unify kernel_physical_mapping_init() API

Hi Hiroyuki-san,

KAMEZAWA Hiroyuki kirjoitti:
> On Wed, 24 Feb 2010 17:04:47 +0200 (EET)
> Pekka J Enberg <penberg@...helsinki.fi> wrote:
> 
>> From: Pekka Enberg <penberg@...helsinki.fi>
>>
>> This patch changes the 32-bit version of kernel_physical_mapping_init() to
>> return the last mapped address like the 64-bit one so that we can unify the
>> call-site in init_memory_mapping().
>>
> 
> I'm sorry if I can't track the logic correctly..

OK. I tried to make it match what x86-64 does.

>> Cc: Yinghai Lu <yinghai@...nel.org>
>> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@...fujitsu.com>
>> Signed-off-by: Pekka Enberg <penberg@...helsinki.fi>
>> ---
>> Note: I have only tested this on VirtualBox which is why I tagged the patch as
>> RFT.
>>
>>   arch/x86/mm/init.c    |    7 -------
>>   arch/x86/mm/init_32.c |    8 +++++---
>>   2 files changed, 5 insertions(+), 10 deletions(-)
>>
>> diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
>> index d406c52..e71c5cb 100644
>> --- a/arch/x86/mm/init.c
>> +++ b/arch/x86/mm/init.c
>> @@ -266,16 +266,9 @@ unsigned long __init_refok init_memory_mapping(unsigned long start,
>>   	if (!after_bootmem)
>>   		find_early_table_space(end, use_pse, use_gbpages);
>>
>> -#ifdef CONFIG_X86_32
>> -	for (i = 0; i < nr_range; i++)
>> -		kernel_physical_mapping_init(mr[i].start, mr[i].end,
>> -					     mr[i].page_size_mask);
>> -	ret = end;
>> -#else /* CONFIG_X86_64 */
>>   	for (i = 0; i < nr_range; i++)
>>   		ret = kernel_physical_mapping_init(mr[i].start, mr[i].end,
>>   						   mr[i].page_size_mask);
>> -#endif
>>
>>   #ifdef CONFIG_X86_32
>>   	early_ioremap_page_table_range_init();
>> diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
>> index 9a0c258..2226f2c 100644
>> --- a/arch/x86/mm/init_32.c
>> +++ b/arch/x86/mm/init_32.c
>> @@ -241,6 +241,7 @@ kernel_physical_mapping_init(unsigned long start,
>>   			     unsigned long page_size_mask)
>>   {
>>   	int use_pse = page_size_mask == (1<<PG_LEVEL_2M);
>> +	unsigned long last_map_addr = end;
>>   	unsigned long start_pfn, end_pfn;
>>   	pgd_t *pgd_base = swapper_pg_dir;
>>   	int pgd_idx, pmd_idx, pte_ofs;
>> @@ -341,9 +342,10 @@ repeat:
>>   					prot = PAGE_KERNEL_EXEC;
>>
>>   				pages_4k++;
>> -				if (mapping_iter == 1)
>> +				if (mapping_iter == 1) {
>>   					set_pte(pte, pfn_pte(pfn, init_prot));
>> -				else
>> +					last_map_addr = (pfn << PAGE_SHIFT) + PAGE_SIZE;
>> +				} else
>>   					set_pte(pte, pfn_pte(pfn, prot));
>>   			}
>>   		}
> 
> Don't we need to update last_map_addr after pages_2m++; ?

Yeah, I missed that part. It probably works fine because we return "end" 
for 2M pages. However, to fix that, I'm not completely sure what 
"last_map_addr" would be. I am thinking

   last_map_addr = (addr & PMD_MASK) + PMD_SIZE;

might do the trick but I'm not that familiar with the code. Ideas?

> And...from the logic,
>   unsigned long last_map_addr = end;
> seems to be always true.
> 
> So, just returning "end" is not good ?

Might be but like I said, I tried to make it match x86-64 so that we 
could eventually try to unify those bits as well.

			Pekka
--
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/