Re: [PATCH v4] kho: kexec-metadata: track previous kernel chain

[Mike Rapoport <rppt@kernel.org>]

Hi Breno,

On Wed, Jan 21, 2026 at 06:50:38AM -0800, Breno Leitao wrote:
> Use Kexec Handover (KHO) to pass the previous kernel's version string
> and the number of kexec reboots since the last cold boot to the next
> kernel, and print it at boot time.
> 
> Example output:
>     [    0.000000] KHO: exec from: 6.19.0-rc4-next-20260107 (count 1)
> 
> Motivation
> ==========
> 
> Bugs that only reproduce when kexecing from specific kernel versions
> are difficult to diagnose. These issues occur when a buggy kernel
> kexecs into a new kernel, with the bug manifesting only in the second
> kernel.
> 
> Recent examples include the following commits:
> 
>  * eb2266312507 ("x86/boot: Fix page table access in 5-level to 4-level paging transition")
>  * 77d48d39e991 ("efistub/tpm: Use ACPI reclaim memory for event log to avoid corruption")
>  * 64b45dd46e15 ("x86/efi: skip memattr table on kexec boot")
> 
> As kexec-based reboots become more common, these version-dependent bugs
> are appearing more frequently. At scale, correlating crashes to the
> previous kernel version is challenging, especially when issues only
> occur in specific transition scenarios.
> 
> Implementation
> ==============
> 
> The kexec metadata is stored as a plain C struct (struct kho_kexec_metadata)
> rather than FDT format, for simplicity and direct field access. It is
> registered via kho_add_subtree() as a separate subtree, keeping it
> independent from the core KHO ABI. This design choice:
> 
>  - Keeps the core KHO ABI minimal and stable
>  - Allows the metadata format to evolve independently
>  - Avoids requiring version bumps for all KHO consumers (LUO, etc.)
>    when the metadata format changes
> 
> The struct kho_metadata contains two fields:
>  - previous_release: The kernel version that initiated the kexec
>  - kexec_count: Number of kexec boots since last cold boot
> 
> On cold boot, kexec_count starts at 0 and increments with each kexec.
> The count helps identify issues that only manifest after multiple
> consecutive kexec reboots.
> 
> Signed-off-by: Breno Leitao <leitao@debian.org>
> Acked-by: SeongJae Park <sj@kernel.org>
> ---
> Changes in v4:
> - Squashed everything in a single commit
> - Moved from FDT to C structs (Pratyush)
> - Usage of subtress intead of FDT directly (Pratyush)
> - Renamed a bunch of variables and functions.
> - Link to v3: https://patch.msgid.link/20260108-kho-v3-0-b1d6b7a89342@debian.org
> 
> Changes in v3:
> - Remove the extra CONFIG for this feature.
> - Reworded some identifiers, properties and printks.
> - Better documented the questions raised during v2.
> - Link to v2: https://patch.msgid.link/20260102-kho-v2-0-1747b1a3a1d6@debian.org
> 
> Changes from v2 to v1 (RFC)
> - Track the number of kexecs since cold boot (Pasha)
> - Change the printk() order compared to KHO
> - Rewording of the commit summary
> - Link to RFC: https://patch.msgid.link/20251230-kho-v1-1-4d795a24da9e@debian.org
> ---
>  include/linux/kho/abi/kexec_handover.h | 29 +++++++++++++++
>  kernel/liveupdate/kexec_handover.c     | 65 ++++++++++++++++++++++++++++++++++
>  2 files changed, 94 insertions(+)
> 
> diff --git a/include/linux/kho/abi/kexec_handover.h b/include/linux/kho/abi/kexec_handover.h
> index 285eda8a36e45..e18022a4e664d 100644
> --- a/include/linux/kho/abi/kexec_handover.h
> +++ b/include/linux/kho/abi/kexec_handover.h
> @@ -11,6 +11,7 @@
>  #define _LINUX_KHO_ABI_KEXEC_HANDOVER_H
>  
>  #include <linux/types.h>
> +#include <linux/utsname.h>
>  
>  /**
>   * DOC: Kexec Handover ABI
> @@ -84,6 +85,34 @@
>  /* The FDT property for sub-FDTs. */
>  #define KHO_FDT_SUB_TREE_PROP_NAME "fdt"
>  
> +/**
> + * DOC: Kexec Metadata ABI
> + *

It would be nice to link it from Documentation/ as well ;-)

> + * The "kexec-metadata" subtree stores optional metadata about the kexec chain.
> + * It is registered via kho_add_subtree(), keeping it independent from the core
> + * KHO ABI. This allows the metadata format to evolve without affecting other
> + * KHO consumers.
> + *
> + * The metadata is stored as a plain C struct rather than FDT format for
> + * simplicity and direct field access.
> + */
> +
> +/**
> + * struct kho_kexec_metadata - Kexec metadata passed between kernels
> + * @previous_release: Kernel version string that initiated the kexec
> + * @kexec_count: Number of kexec boots since last cold boot
> + *
> + * This structure is preserved across kexec and allows the new kernel to
> + * identify which kernel it was booted from and how many kexec reboots
> + * have occurred.
> + */
> +struct kho_kexec_metadata {
> +	char previous_release[__NEW_UTS_LEN + 1];
> +	u32 kexec_count;
> +} __packed;
> +
> +#define KHO_METADATA_NODE_NAME "kexec-metadata"
> +
>  /**
>   * DOC: Kexec Handover ABI for vmalloc Preservation
>   *
> diff --git a/kernel/liveupdate/kexec_handover.c b/kernel/liveupdate/kexec_handover.c

...

>  static __init int kho_init(void)
>  {
>  	const void *fdt = kho_get_fdt();
> @@ -1357,6 +1413,15 @@ static __init int kho_init(void)
>  	if (err)
>  		goto err_free_fdt;
>  
> +	if (fdt)
> +		kho_process_kexec_metadata();

Can't we move it into the existing if (fdt) below?
 
> +
> +	/* Populate kexec metadata for the possible next kexec */
> +	err = kho_populate_kexec_metadata();
> +	if (err)
> +		pr_warn("failed to initialize kexec-metadata subtree: %d\n",
> +			err);

Please follow if (err) goto err_ pattern.

kho_populate_kexec_metadata() failure essentially means that we failed to
allocate memory. This shouldn't happen that early in boot, but if it did,
then something is utterly wrong.

> +
>  	if (fdt) {
>  		kho_in_debugfs_init(&kho_in.dbg, fdt);
>  		return 0;

-- 
Sincerely yours,
Mike.