All of lore.kernel.org
 help / color / mirror / Atom feed
From: Steven Price <steven.price@arm.com>
To: Mark Rutland <mark.rutland@arm.com>
Cc: x86@kernel.org, "Arnd Bergmann" <arnd@arndb.de>,
	"Ard Biesheuvel" <ard.biesheuvel@linaro.org>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Catalin Marinas" <catalin.marinas@arm.com>,
	"Dave Hansen" <dave.hansen@linux.intel.com>,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	"Jérôme Glisse" <jglisse@redhat.com>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"Andy Lutomirski" <luto@kernel.org>,
	"H. Peter Anvin" <hpa@zytor.com>,
	"James Morse" <james.morse@arm.com>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Will Deacon" <will@kernel.org>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	linux-arm-kernel@lists.infradead.org, "Liang,
	Kan" <kan.liang@linux.intel.com>
Subject: Re: [PATCH v9 19/21] mm: Add generic ptdump
Date: Wed, 24 Jul 2019 17:36:45 +0100	[thread overview]
Message-ID: <ee707646-0196-63bb-45cc-6b949ae9530e@arm.com> (raw)
In-Reply-To: <20190723095747.GB8085@lakrids.cambridge.arm.com>

On 23/07/2019 10:57, Mark Rutland wrote:
> On Mon, Jul 22, 2019 at 04:42:08PM +0100, Steven Price wrote:
>> Add a generic version of page table dumping that architectures can
>> opt-in to
>>
>> Signed-off-by: Steven Price <steven.price@arm.com>
> 
> [...]
> 
>> +#ifdef CONFIG_KASAN
>> +/*
>> + * This is an optimization for KASAN=y case. Since all kasan page tables
>> + * eventually point to the kasan_early_shadow_page we could call note_page()
>> + * right away without walking through lower level page tables. This saves
>> + * us dozens of seconds (minutes for 5-level config) while checking for
>> + * W+X mapping or reading kernel_page_tables debugfs file.
>> + */
>> +static inline bool kasan_page_table(struct ptdump_state *st, void *pt,
>> +				    unsigned long addr)
>> +{
>> +	if (__pa(pt) == __pa(kasan_early_shadow_pmd) ||
>> +#ifdef CONFIG_X86
>> +	    (pgtable_l5_enabled() &&
>> +			__pa(pt) == __pa(kasan_early_shadow_p4d)) ||
>> +#endif
>> +	    __pa(pt) == __pa(kasan_early_shadow_pud)) {
>> +		st->note_page(st, addr, 5, pte_val(kasan_early_shadow_pte[0]));
>> +		return true;
>> +	}
>> +	return false;
> 
> Having you tried this with CONFIG_DEBUG_VIRTUAL?
> 
> The kasan_early_shadow_pmd is a kernel object rather than a linear map
> object, so you should use __pa_symbol for that.

Thanks for pointing that out - it is indeed broken on arm64. This was
moved from x86 where CONFIG_DEBUG_VIRTUAL doesn't seem to pick this up.
There is actually a problem here that 'pt' might not be in the linear
map (so __pa(pt) barfs on arm64 as well as kasan_early_shadow_p?d).

It looks like having the comparisons of the form "pt ==
lm_alias(kasan_early_shadow_p?d)" is probably best.

> It's a bit horrid to have to test multiple levels in one function; can't
> we check the relevant level inline in each of the test_p?d funcs?
> 
> They're optional anyway, so they only need to be defined for
> CONFIG_KASAN.

Good point - removing the test_p?d callbacks when !CONFIG_KASAN
simplifies the code.

Thanks,

Steve

WARNING: multiple messages have this Message-ID (diff)
From: Steven Price <steven.price@arm.com>
To: Mark Rutland <mark.rutland@arm.com>
Cc: "Dave Hansen" <dave.hansen@linux.intel.com>,
	"Arnd Bergmann" <arnd@arndb.de>,
	"Ard Biesheuvel" <ard.biesheuvel@linaro.org>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Catalin Marinas" <catalin.marinas@arm.com>,
	x86@kernel.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	"Jérôme Glisse" <jglisse@redhat.com>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"Andy Lutomirski" <luto@kernel.org>,
	"H. Peter Anvin" <hpa@zytor.com>,
	"James Morse" <james.morse@arm.com>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Will Deacon" <will@kernel.org>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	linux-arm-kernel@lists.infradead.org, "Liang,
	Kan" <kan.liang@linux.intel.com>
Subject: Re: [PATCH v9 19/21] mm: Add generic ptdump
Date: Wed, 24 Jul 2019 17:36:45 +0100	[thread overview]
Message-ID: <ee707646-0196-63bb-45cc-6b949ae9530e@arm.com> (raw)
In-Reply-To: <20190723095747.GB8085@lakrids.cambridge.arm.com>

On 23/07/2019 10:57, Mark Rutland wrote:
> On Mon, Jul 22, 2019 at 04:42:08PM +0100, Steven Price wrote:
>> Add a generic version of page table dumping that architectures can
>> opt-in to
>>
>> Signed-off-by: Steven Price <steven.price@arm.com>
> 
> [...]
> 
>> +#ifdef CONFIG_KASAN
>> +/*
>> + * This is an optimization for KASAN=y case. Since all kasan page tables
>> + * eventually point to the kasan_early_shadow_page we could call note_page()
>> + * right away without walking through lower level page tables. This saves
>> + * us dozens of seconds (minutes for 5-level config) while checking for
>> + * W+X mapping or reading kernel_page_tables debugfs file.
>> + */
>> +static inline bool kasan_page_table(struct ptdump_state *st, void *pt,
>> +				    unsigned long addr)
>> +{
>> +	if (__pa(pt) == __pa(kasan_early_shadow_pmd) ||
>> +#ifdef CONFIG_X86
>> +	    (pgtable_l5_enabled() &&
>> +			__pa(pt) == __pa(kasan_early_shadow_p4d)) ||
>> +#endif
>> +	    __pa(pt) == __pa(kasan_early_shadow_pud)) {
>> +		st->note_page(st, addr, 5, pte_val(kasan_early_shadow_pte[0]));
>> +		return true;
>> +	}
>> +	return false;
> 
> Having you tried this with CONFIG_DEBUG_VIRTUAL?
> 
> The kasan_early_shadow_pmd is a kernel object rather than a linear map
> object, so you should use __pa_symbol for that.

Thanks for pointing that out - it is indeed broken on arm64. This was
moved from x86 where CONFIG_DEBUG_VIRTUAL doesn't seem to pick this up.
There is actually a problem here that 'pt' might not be in the linear
map (so __pa(pt) barfs on arm64 as well as kasan_early_shadow_p?d).

It looks like having the comparisons of the form "pt ==
lm_alias(kasan_early_shadow_p?d)" is probably best.

> It's a bit horrid to have to test multiple levels in one function; can't
> we check the relevant level inline in each of the test_p?d funcs?
> 
> They're optional anyway, so they only need to be defined for
> CONFIG_KASAN.

Good point - removing the test_p?d callbacks when !CONFIG_KASAN
simplifies the code.

Thanks,

Steve

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  reply	other threads:[~2019-07-24 16:36 UTC|newest]

Thread overview: 129+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-22 15:41 [PATCH v9 00/21] Generic page walk and ptdump Steven Price
2019-07-22 15:41 ` Steven Price
2019-07-22 15:41 ` [PATCH v9 01/21] arc: mm: Add p?d_leaf() definitions Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 02/21] arm: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 03/21] arm64: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 04/21] mips: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 21:47   ` Paul Burton
2019-07-22 21:47     ` Paul Burton
2019-07-24 13:03     ` Steven Price
2019-07-24 13:03       ` Steven Price
2019-07-22 15:41 ` [PATCH v9 05/21] powerpc: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 06/21] riscv: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 07/21] s390: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 08/21] sparc: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 09/21] x86: " Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-22 15:41 ` [PATCH v9 10/21] mm: Add generic p?d_leaf() macros Steven Price
2019-07-22 15:41   ` Steven Price
2019-07-23  9:41   ` Mark Rutland
2019-07-23  9:41     ` Mark Rutland
2019-07-24 13:48     ` Steven Price
2019-07-24 13:48       ` Steven Price
2019-07-28 11:44     ` Anshuman Khandual
2019-07-28 11:44       ` Anshuman Khandual
2019-07-29 11:38       ` Steven Price
2019-07-29 11:38         ` Steven Price
2019-08-01  6:09         ` Anshuman Khandual
2019-08-01  6:09           ` Anshuman Khandual
2019-08-01 12:22           ` Steven Price
2019-08-01 12:22             ` Steven Price
2019-07-29 12:50       ` Mark Rutland
2019-07-29 12:50         ` Mark Rutland
2019-08-01  6:13         ` Anshuman Khandual
2019-08-01  6:13           ` Anshuman Khandual
2019-07-22 15:42 ` [PATCH v9 11/21] mm: pagewalk: Add p4d_entry() and pgd_entry() Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-23 10:14   ` Mark Rutland
2019-07-23 10:14     ` Mark Rutland
2019-07-24 13:53     ` Steven Price
2019-07-24 13:53       ` Steven Price
2019-07-24 14:09       ` Mark Rutland
2019-07-24 14:09         ` Mark Rutland
2019-07-28 12:33   ` Anshuman Khandual
2019-07-28 12:33     ` Anshuman Khandual
2019-07-29 12:17     ` Steven Price
2019-07-29 12:17       ` Steven Price
2019-07-22 15:42 ` [PATCH v9 12/21] mm: pagewalk: Allow walking without vma Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-28 14:20   ` Anshuman Khandual
2019-07-28 14:20     ` Anshuman Khandual
2019-07-29 12:29     ` Steven Price
2019-07-29 12:29       ` Steven Price
2019-08-01  6:41       ` Anshuman Khandual
2019-08-01  6:41         ` Anshuman Khandual
2019-07-22 15:42 ` [PATCH v9 13/21] mm: pagewalk: Add test_p?d callbacks Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-28 13:41   ` Anshuman Khandual
2019-07-28 13:41     ` Anshuman Khandual
2019-07-29 12:34     ` Steven Price
2019-07-29 12:34       ` Steven Price
2019-07-22 15:42 ` [PATCH v9 14/21] x86: mm: Don't display pages which aren't present in debugfs Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 15/21] x86: mm: Point to struct seq_file from struct pg_state Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 16/21] x86: mm+efi: Convert ptdump_walk_pgd_level() to take a mm_struct Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 17/21] x86: mm: Convert ptdump_walk_pgd_level_debugfs() to take an mm_struct Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 18/21] x86: mm: Convert ptdump_walk_pgd_level_core() " Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 19/21] mm: Add generic ptdump Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-23  9:57   ` Mark Rutland
2019-07-23  9:57     ` Mark Rutland
2019-07-24 16:36     ` Steven Price [this message]
2019-07-24 16:36       ` Steven Price
2019-07-29  2:59   ` Anshuman Khandual
2019-07-29  2:59     ` Anshuman Khandual
2019-07-29 13:56     ` Steven Price
2019-07-29 13:56       ` Steven Price
2019-07-22 15:42 ` [PATCH v9 20/21] x86: mm: Convert dump_pagetables to use walk_page_range Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-22 15:42 ` [PATCH v9 21/21] arm64: mm: Convert mm/dump.c to use walk_page_range() Steven Price
2019-07-22 15:42   ` Steven Price
2019-07-23  6:39 ` [PATCH v9 00/21] Generic page walk and ptdump Anshuman Khandual
2019-07-23  6:39   ` Anshuman Khandual
2019-07-24 13:35   ` Steven Price
2019-07-24 13:35     ` Steven Price
2019-07-25  9:09     ` Anshuman Khandual
2019-07-25  9:09       ` Anshuman Khandual
2019-07-25  9:30       ` Will Deacon
2019-07-25  9:30         ` Will Deacon
2019-07-26  6:03         ` Anshuman Khandual
2019-07-26  6:03           ` Anshuman Khandual
2019-07-25 10:15       ` Steven Price
2019-07-25 10:15         ` Steven Price
2019-07-23 10:16 ` Mark Rutland
2019-07-23 10:16   ` Mark Rutland
2019-07-24 13:35   ` Steven Price
2019-07-24 13:35     ` Steven Price
2019-07-24 13:57     ` Thomas Gleixner
2019-07-24 13:57       ` Thomas Gleixner
2019-07-24 14:07       ` Mark Rutland
2019-07-24 14:07         ` Mark Rutland
2019-07-24 14:18       ` Steven Price
2019-07-24 14:18         ` Steven Price
2019-07-24 14:37         ` Thomas Gleixner
2019-07-24 14:37           ` Thomas Gleixner
2019-07-28 11:20 ` Anshuman Khandual
2019-07-28 11:20   ` Anshuman Khandual
2019-07-29 11:32   ` Steven Price
2019-07-29 11:32     ` Steven Price
2019-07-31  9:27     ` Sven Schnelle
2019-07-31  9:27       ` Sven Schnelle
2019-07-31 11:18       ` Steven Price
2019-07-31 11:18         ` Steven Price

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=ee707646-0196-63bb-45cc-6b949ae9530e@arm.com \
    --to=steven.price@arm.com \
    --cc=akpm@linux-foundation.org \
    --cc=ard.biesheuvel@linaro.org \
    --cc=arnd@arndb.de \
    --cc=bp@alien8.de \
    --cc=catalin.marinas@arm.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=hpa@zytor.com \
    --cc=james.morse@arm.com \
    --cc=jglisse@redhat.com \
    --cc=kan.liang@linux.intel.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=luto@kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=will@kernel.org \
    --cc=x86@kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.