Use monotonic clock for sequence numbering

[ZoltanFridrich]

Fixes the following error found by static analysis:

Error: Y2K38_SAFETY (CWE-197):
libkcapi-1.3.1/lib/kcapi-kernel-if.c:658: store_truncates_time_t: A "time_t" value is stored in an integer with too few bits to accommodate it.  The expression "time(NULL)" is cast to "__u32".
#  656|   	req.n.nlmsg_flags = NLM_F_REQUEST;
#  657|   	req.n.nlmsg_type = CRYPTO_MSG_GETALG;
#  658|-> 	req.n.nlmsg_seq = (__u32)time(NULL);
#  659|   
#  660|   	if (drivername)

[smuellerDD]

Am Dienstag, 12. März 2024, 17:19:16 CET schrieb Zoltán Fridrich: Hi Zoltán,

Fixes the following error found by static analysis:

Thanks for reporting, but before I apply that, can you please tell me what your static analyzer complains about? Is it that time() may return -1? If so, why do you think we need to replace time() in general wit clock_gettime? Thanks

``` Error: Y2K38_SAFETY (CWE-197): libkcapi-1.3.1/lib/kcapi-kernel-if.c:658: store_truncates_time_t: A "time_t" value is stored in an integer with too few bits to accommodate it. The expression "time(NULL)" is cast to "__u32". # 656| req.n.nlmsg_flags = NLM_F_REQUEST; # 657| req.n.nlmsg_type = CRYPTO_MSG_GETALG; # 658|-> req.n.nlmsg_seq = (__u32)time(NULL); # 659| # 660| if (drivername) ``` You can view, comment on, or merge this pull request online at: #184 -- Commit Summary -- * Use monotonic clock for sequence numbering -- File Changes -- M lib/kcapi-kernel-if.c (9) -- Patch Links -- https://github.com/smuellerDD/libkcapi/pull/184.patch https://github.com/smuellerDD/libkcapi/pull/184.diff

Ciao Stephan

[ZoltanFridrich]

For some reason I thought that when retyping a larger type to smaller type an overflow would cause the count to stop. But of course it will continue from zero. This is a false positive. Please disregard this MR.