linux-kernel - Re: [PATCH v2] staging: wilc1000: Fix sparse warnings incorrect type assignment

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]

Date:   Wed, 8 Mar 2017 09:22:06 +0300
From:   Dan Carpenter <dan.carpenter@...cle.com>
To:     Andrea Ghittino <aghittino@...il.com>
Cc:     aditya.shankar@...rochip.com, ganesh.krishna@...rochip.com,
        gregkh@...uxfoundation.org, linux-wireless@...r.kernel.org,
        devel@...verdev.osuosl.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] staging: wilc1000: Fix sparse warnings incorrect type
 assignment

I think this change is buggy.

On Tue, Mar 07, 2017 at 10:36:53PM +0100, Andrea Ghittino wrote:
> Fixed sparse warnings related to the conversion of le16 and le32 to u16 and u32, during the update of internal structures
> 
> Fixed sparse warnings:
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2011:52: warning: incorrect type in assignment (different base types)
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2011:52:    expected unsigned short [unsigned] [assigned] [usertype] ht_ext_params
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2011:52:    got restricted __le16 const [usertype] extended_ht_cap_info
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2012:51: warning: incorrect type in assignment (different base types)
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2012:51:    expected unsigned int [unsigned] [assigned] [usertype] ht_tx_bf_cap
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2012:51:    got restricted __le32 const [usertype] tx_BF_cap_info
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2078:51: warning: incorrect type in assignment (different base types)
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2078:51:    expected unsigned short [unsigned] [assigned] [usertype] ht_capa_info
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2078:51:    got restricted __le16 const [usertype] cap_info
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2083:52: warning: incorrect type in assignment (different base types)
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2083:52:    expected unsigned short [unsigned] [assigned] [usertype] ht_ext_params
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2083:52:    got restricted __le16 const [usertype] extended_ht_cap_info
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2084:51: warning: incorrect type in assignment (different base types)
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2084:51:    expected unsigned int [unsigned] [assigned] [usertype] ht_tx_bf_cap
> drivers/staging/wilc1000//wilc_wfi_cfgoperations.c:2084:51:    got restricted __le32 const [usertype] tx_BF_cap_info
> 

What I want in a changelog is:  "We're copying data that's little endian
from (some place) to where ever and then we (whatever) do math using the
variable so it needs to be CPU endian.  Presumably this wasn't caught
in testing because it was only used on x86 or other little endian
systems."

In this case we're copying little endian data and then sending it
directly back to some place which requires little endian data so
converting it is a bug.

regards,
dan carpenter