linux-kernel - Re: [PATCH v2 2/2] kexec: history: track kexec boot counter

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

Message-ID: <rsygqnm3qahkfoqo6usrz3jhpya3c3rqebz6rotl33dsx3rsbo@2v4yeabi7fda>
Date: Fri, 2 Jan 2026 07:24:54 -0800
From: Breno Leitao <leitao@...ian.org>
To: Usama Arif <usamaarif642@...il.com>
Cc: Alexander Graf <graf@...zon.com>, Mike Rapoport <rppt@...nel.org>, 
	Pasha Tatashin <pasha.tatashin@...een.com>, Pratyush Yadav <pratyush@...nel.org>, 
	linux-kernel@...r.kernel.org, kexec@...ts.infradead.org, linux-mm@...ck.org, rmikey@...a.com, 
	clm@...com, riel@...riel.com, kernel-team@...a.com
Subject: Re: [PATCH v2 2/2] kexec: history: track kexec boot counter

Hello Usama,

On Fri, Jan 02, 2026 at 06:09:57PM +0300, Usama Arif wrote:
> 
> On 02/01/2026 17:53, Breno Leitao wrote:
> > Track and display the number of kexec boots since the last cold reboot
> > when CONFIG_KEXEC_HISTORY is enabled.
> > 
> > This extends the previous kernel release tracking feature by adding
> > a counter that increments with each kexec boot. The counter provides
> > visibility into the kexec chain depth, which is useful for understanding
> > boot history in production environments.
> > 
> > Add a new property, "kexec-count" in KHO FDT alongside the existing
> > "previous-release" property. The counter is:
> > 
> >  - Initialized to 0 when kho_in is instantiated.
> >  - Incremented by 1 on each subsequent kexec.
> >  - Printed alongside the previous kernel release version.
> > 
> > The counter is stored as a 32-bit unsigned integer in FDT format and is
> > only active when CONFIG_KEXEC_HISTORY is enabled.
> > 
> > Signed-off-by: Breno Leitao <leitao@...ian.org>
> > Suggested-by: Pasha Tatashin <pasha.tatashin@...een.com>
> > ---
> >  kernel/liveupdate/kexec_handover.c | 25 +++++++++++++++++++++++--
> >  1 file changed, 23 insertions(+), 2 deletions(-)
> > 
> > diff --git a/kernel/liveupdate/kexec_handover.c b/kernel/liveupdate/kexec_handover.c
> > index 06d99627bb3c..fe5a2c5c4c86 100644
> > --- a/kernel/liveupdate/kexec_handover.c
> > +++ b/kernel/liveupdate/kexec_handover.c
> > @@ -38,6 +38,7 @@
> >  #define PROP_PRESERVED_MEMORY_MAP "preserved-memory-map"
> >  #define PROP_SUB_FDT "fdt"
> >  #define PROP_PREVIOUS_RELEASE "previous-release"
> > +#define PROP_KEXEC_COUNT "kexec-count"
> >  
> >  #define KHO_PAGE_MAGIC 0x4b484f50U /* ASCII for 'KHOP' */
> >  
> > @@ -1257,6 +1258,7 @@ struct kho_in {
> >  	phys_addr_t scratch_phys;
> >  #ifdef CONFIG_KEXEC_HISTORY
> >  	char previous_release[__NEW_UTS_LEN + 1];
> > +	u32 kexec_count;
> >  #endif
> >  	struct kho_debugfs dbg;
> >  };
> > @@ -1330,6 +1332,9 @@ static __init int kho_out_fdt_setup(void)
> >  	void *root = kho_out.fdt;
> >  	u64 empty_mem_map = 0;
> >  	int err;
> > +#ifdef CONFIG_KEXEC_HISTORY
> > +	u32 kexec_count;
> > +#endif
> >  
> >  	err = fdt_create(root, PAGE_SIZE);
> >  	err |= fdt_finish_reservemap(root);
> > @@ -1340,6 +1345,10 @@ static __init int kho_out_fdt_setup(void)
> >  #ifdef CONFIG_KEXEC_HISTORY
> >  	err |= fdt_property_string(root, PROP_PREVIOUS_RELEASE,
> >  				   init_uts_ns.name.release);
> > +	/* kho_in.kexec_count is set to 0 on cold boot */
> > +	kexec_count = cpu_to_fdt32(kho_in.kexec_count + 1);
> 
> Should this be kexec_count = cpu_to_fdt32(kho_in.kexec_count) + 1; ?

I don't think so. Basically we want to increment the counter in native
endianess and then convert to the Big Endian (FDT/DT are big endian
AFAIK) and then store (line below) to the DT.

If we convert to big endian and then sum 1, it will mess with the
result. 

> +    err |= fdt_property(root, PROP_KEXEC_COUNT, &kexec_count,
> > +                sizeof(kexec_count));

Thanks for the review,
--breno