linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Anshuman Khandual <anshuman.khandual@arm.com>
To: "Ritesh Harjani (IBM)" <ritesh.list@gmail.com>,
	linux-arm-kernel@lists.infradead.org
Cc: Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will@kernel.org>, Mark Brown <broonie@kernel.org>,
	Ryan Roberts <ryan.roberts@arm.com>,
	Mark Rutland <mark.rutland@arm.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [RFC 3/3] arm64/ptdump: Add ARM64_PTDUMP_CONSOLE
Date: Thu, 28 Aug 2025 07:52:51 +0530	[thread overview]
Message-ID: <7302615c-e190-43ac-bc29-e82d7b048403@arm.com> (raw)
In-Reply-To: <87bjo2qe5w.fsf@gmail.com>

On 26/08/25 9:16 PM, Ritesh Harjani (IBM) wrote:
> Anshuman Khandual <anshuman.khandual@arm.com> writes:
> 
>> Enable early kernel page table dump for debug purpose when required via new
>> config ARM64_PDUMP_CONSOLE. This calls existing ptdump_walk() early on just
>> after ptdump has been initialized with ptdump_init().
> 
> I happen to stumble upon this while looking for something else related
> to ptdump and was curious to understand where this will be really
> useful? 
> 
> So instead of dumping it via cat /sys/kernel/debug/kernel_page_tables,
> this will dump at early boot during arch setup and before start_kernel().

Right, primarily before vmalloc() space gets crowded. Also this provides
an opportunity to do a diff between early boot and after boot kernel page
table states.
> 
> I was curious, since this anyway gets enabled only in debug kernels.
> There we can always just boot with minimal busybox image which can jump
> to shell quickly and dump the kernel page tables, correct?

Here the kernel page table dump could happen earlier than that as well.
> 
> Also is ARM64_PTDUMP_CONSOLE config option added on purpose? A kernel cmdline
> like early_ptdump=yes|1|true could come much handy, right?

Currently this is just for arm64 platform but could be enabled in general
for other platforms as well. Yes, early_ptdump=yes|1|true will be useful
as well. ARM64_PTDUMP_CONSOLE just build guards the additional code. But
if required cmdline option "early_ptdump=" could just provide the runtime
switch and then this could always be built enabled on CONFIG_PTDUMP.
> 
> (Since I am fixing few issues on powerpc ptdump - hence was just curious
> to know whether this can come useful for me too or not :) )

This feature could be extended in general to other platforms. Let me know
if you find this useful for powerpc.
> 
> Thanks!
> -ritesh
> 
>>
>> Suggested-by: Ryan Roberts <ryan.roberts@arm.com>
>> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
>> ---
>>  arch/arm64/Kconfig.debug        | 12 ++++++++++++
>>  arch/arm64/include/asm/ptdump.h |  7 +++++++
>>  arch/arm64/kernel/setup.c       |  1 +
>>  arch/arm64/mm/ptdump.c          |  7 +++++++
>>  4 files changed, 27 insertions(+)
>>
>> diff --git a/arch/arm64/Kconfig.debug b/arch/arm64/Kconfig.debug
>> index 265c4461031f..0f8af0dd0f4c 100644
>> --- a/arch/arm64/Kconfig.debug
>> +++ b/arch/arm64/Kconfig.debug
>> @@ -20,4 +20,16 @@ config ARM64_RELOC_TEST
>>  	depends on m
>>  	tristate "Relocation testing module"
>>  
>> +config ARM64_PTDUMP_CONSOLE
>> +	bool "Dump early kernel page table"
>> +	depends on DEBUG_KERNEL
>> +	depends on ARCH_HAS_PTDUMP
>> +	select PTDUMP
>> +	help
>> +	  Enable this option to dump early kernel page table entries during
>> +	  boot using the PTDUMP framework. This helps in examining kernel's
>> +	  page table mapping entries and their attributes etc.
>> +
>> +	  If in doubt, say N.
>> +
>>  source "drivers/hwtracing/coresight/Kconfig"
>> diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
>> index 27e774134e7f..81dc53ca9643 100644
>> --- a/arch/arm64/include/asm/ptdump.h
>> +++ b/arch/arm64/include/asm/ptdump.h
>> @@ -74,8 +74,15 @@ void __init ptdump_debugfs_register(struct ptdump_info *info, const char *name);
>>  static inline void ptdump_debugfs_register(struct ptdump_info *info,
>>  					   const char *name) { }
>>  #endif /* CONFIG_PTDUMP_DEBUGFS */
>> +
>> +#ifdef CONFIG_ARM64_PTDUMP_CONSOLE
>> +void __init arm64_kernel_pgtable_dump(void);
>> +#else
>> +static inline void __init arm64_kernel_pgtable_dump(void) { }
>> +#endif /* CONFIG_ARM64_PTDUMP_CONSOLE */
>>  #else
>>  static inline void __init ptdump_init(void) { }
>> +static inline void __init arm64_kernel_pgtable_dump(void) { }
>>  static inline void note_page(struct ptdump_state *pt_st, unsigned long addr,
>>  			     int level, pteval_t val) { }
>>  static inline void note_page_pte(struct ptdump_state *st, unsigned long addr, pte_t pte) { }
>> diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c
>> index 0a3812c8e177..86bf7607d304 100644
>> --- a/arch/arm64/kernel/setup.c
>> +++ b/arch/arm64/kernel/setup.c
>> @@ -361,6 +361,7 @@ void __init __no_sanitize_address setup_arch(char **cmdline_p)
>>  	init_bootcpu_ops();
>>  	smp_init_cpus();
>>  	smp_build_mpidr_hash();
>> +	arm64_kernel_pgtable_dump();
>>  
>>  #ifdef CONFIG_ARM64_SW_TTBR0_PAN
>>  	/*
>> diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
>> index c78e6b496dea..f6d22462add6 100644
>> --- a/arch/arm64/mm/ptdump.c
>> +++ b/arch/arm64/mm/ptdump.c
>> @@ -407,6 +407,13 @@ void __init ptdump_init(void)
>>  	ptdump_initialize();
>>  }
>>  
>> +#ifdef CONFIG_ARM64_PTDUMP_CONSOLE
>> +void __init arm64_kernel_pgtable_dump(void)
>> +{
>> +	ptdump_walk(CONSOLE, &kernel_ptdump_info);
>> +}
>> +#endif
>> +
>>  static int __init ptdump_debugfs_init(void)
>>  {
>>  	ptdump_debugfs_register(&kernel_ptdump_info, "kernel_page_tables");
>> -- 
>> 2.25.1



  reply	other threads:[~2025-08-28  2:25 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-18  9:14 [RFC 0/3] arm64/ptdump: Add ARM64_PTDUMP_CONSOLE Anshuman Khandual
2025-08-18  9:14 ` [RFC 1/3] arm64/ptdump: Re-organize ptdump_init() Anshuman Khandual
2025-08-18  9:14 ` [RFC 2/3] arm64/ptdump: Enable console output in pt_dump_seq_[printf|puts]() Anshuman Khandual
2025-08-18  9:14 ` [RFC 3/3] arm64/ptdump: Add ARM64_PTDUMP_CONSOLE Anshuman Khandual
2025-08-26 15:46   ` Ritesh Harjani
2025-08-28  2:22     ` Anshuman Khandual [this message]
2025-08-29  3:42       ` Ritesh Harjani
2025-08-29  6:12         ` Anshuman Khandual

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=7302615c-e190-43ac-bc29-e82d7b048403@arm.com \
    --to=anshuman.khandual@arm.com \
    --cc=broonie@kernel.org \
    --cc=catalin.marinas@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=ritesh.list@gmail.com \
    --cc=ryan.roberts@arm.com \
    --cc=will@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 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).