CVE-2024-38616

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> wifi: carl9170: re-fix fortified-memset warning<br /> <br /> The carl9170_tx_release() function sometimes triggers a fortified-memset<br /> warning in my randconfig builds:<br /> <br /> In file included from include/linux/string.h:254,<br /> from drivers/net/wireless/ath/carl9170/tx.c:40:<br /> In function &amp;#39;fortify_memset_chk&amp;#39;,<br /> inlined from &amp;#39;carl9170_tx_release&amp;#39; at drivers/net/wireless/ath/carl9170/tx.c:283:2,<br /> inlined from &amp;#39;kref_put&amp;#39; at include/linux/kref.h:65:3,<br /> inlined from &amp;#39;carl9170_tx_put_skb&amp;#39; at drivers/net/wireless/ath/carl9170/tx.c:342:9:<br /> include/linux/fortify-string.h:493:25: error: call to &amp;#39;__write_overflow_field&amp;#39; declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror=attribute-warning]<br /> 493 | __write_overflow_field(p_size_field, size);<br /> <br /> Kees previously tried to avoid this by using memset_after(), but it seems<br /> this does not fully address the problem. I noticed that the memset_after()<br /> here is done on a different part of the union (status) than the original<br /> cast was from (rate_driver_data), which may confuse the compiler.<br /> <br /> Unfortunately, the memset_after() trick does not work on driver_rates[]<br /> because that is part of an anonymous struct, and I could not get<br /> struct_group() to do this either. Using two separate memset() calls<br /> on the two members does address the warning though.