linux-hardening - Re: [PATCH v3][next] acpi: nfit: intel: avoid multiple -Wflex-array-member-not-at-end warnings

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: <e0adad17-5d4f-4309-9975-81971597da65@embeddedor.com>
Date: Wed, 25 Jun 2025 11:31:33 -0600
From: "Gustavo A. R. Silva" <gustavo@...eddedor.com>
To: Kees Cook <kees@...nel.org>, "Gustavo A. R. Silva" <gustavoars@...nel.org>
Cc: Dan Williams <dan.j.williams@...el.com>,
 Vishal Verma <vishal.l.verma@...el.com>, Dave Jiang <dave.jiang@...el.com>,
 Ira Weiny <ira.weiny@...el.com>, "Rafael J. Wysocki" <rafael@...nel.org>,
 Len Brown <lenb@...nel.org>, nvdimm@...ts.linux.dev,
 linux-acpi@...r.kernel.org, linux-kernel@...r.kernel.org,
 linux-hardening@...r.kernel.org
Subject: Re: [PATCH v3][next] acpi: nfit: intel: avoid multiple
 -Wflex-array-member-not-at-end warnings



On 25/06/25 10:56, Kees Cook wrote:
> On Wed, Jun 11, 2025 at 01:52:41PM -0600, Gustavo A. R. Silva wrote:
>> -Wflex-array-member-not-at-end was introduced in GCC-14, and we are
>> getting ready to enable it, globally.
>>
>> Refactor multiple structs that contain flexible-array members in the
>> middle by replacing them with unions.
>>
>> These changes preserve the memory layout while effectively adjusting
>> it so that the flexible-array member is always treated as the last
>> member.
>>
>> With these changes, fix a dozen instances of the following type of
>> warning:
>>
>> drivers/acpi/nfit/intel.c:692:35: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
>>
>> Signed-off-by: Gustavo A. R. Silva <gustavoars@...nel.org>
>> ---
>> Changes in v3:
>>   - Use union instead of DEFINE_RAW_FLEX().
> 
> I think your TRAILING_OVERLAP macro[1] is perfect here. I'll try to get that
> landed for the next rc. Can you double-check that this works correctly
> in these cases?

Absolutely. If people prefer that route I'm happy to wait for the helper to
land in linus' tree.

> 
>> @@ -55,9 +55,16 @@ static unsigned long intel_security_flags(struct nvdimm *nvdimm,
>>   {
>>   	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
>>   	unsigned long security_flags = 0;
>> -	struct {
>> +	/*
>> +	 * This effectively creates a union between the flexible-array member
>> +	 * and any members after _offset_to_fam.
>> +	 */
>> +	union {
>>   		struct nd_cmd_pkg pkg;
>> -		struct nd_intel_get_security_state cmd;
>> +		struct {
>> +			u8 _offset_to_fam[offsetof(struct nd_cmd_pkg, nd_payload)];
>> +			struct nd_intel_get_security_state cmd;
>> +		};
>>   	} nd_cmd = {
>>   		.pkg = {
>>   			.nd_command = NVDIMM_INTEL_GET_SECURITY_STATE,
> 
> I think it would be a pretty small and direct replacement:
> 
> 	TRAILING_OVERLAP(struct nd_cmd_pkg, pkg, nd_payload,
> 			 struct nd_intel_get_security_state cmd;
> 	) nd_cmd = {
> 		...

Yes, this works. Hopefully, maintainers will comment on this and let us
know what they prefer. :)

Thanks!
-Gustavo

> 
> -Kees
> 
> [1] https://web.git.kernel.org/pub/scm/linux/kernel/git/kees/linux.git/commit/?h=for-next/kspp&id=29bb79e9dbf1ba100125e39deb7147acd490903f
>