drivers/snmp-ups.c: usmAESPrivProtocol is not an array to properly use a sizeof()

[jimklimov]

snmp-ups.c: In function 'nut_snmp_init':
snmp-ups.c:599:66: error: division 'sizeof (oid * {aka long unsigned int *}) / sizeof (oid {aka long unsigned int})' does not compute the number of array elements [-Werror=sizeof-pointer-div]
  599 |    g_snmp_sess.securityPrivProtoLen =  sizeof(usmAESPrivProtocol)/sizeof(oid);
      |                                                                  ^
In file included from /usr/include/net-snmp/snmpv3_api.h:27,
                 from /usr/include/net-snmp/net-snmp-includes.h:78,
                 from snmp-ups.h:79,
                 from snmp-ups.c:39:
/usr/include/net-snmp/library/transform_oids.h:50:26: note: first 'sizeof' operand was declared here
   50 | NETSNMP_IMPORT oid      *usmAES128PrivProtocol; /* backwards compat */
      |                          ^~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make[1]: *** [Makefile:1953: snmp_ups-snmp-ups.o] Error 1

In my recent fight against warnings in NUT codebase, augmented by #844 and following from #823, this issue remains the only one I don't have an idea about solving. Help would be very much welcome to make a lot of additional Travis CI test-cases green.

[lgtm-com]

This pull request introduces 1 alert when merging 7bcaa25 into 8c1ad08 - view on LGTM.com

new alerts:

• 1 for FIXME comment

[lgtm-com]

This pull request introduces 1 alert when merging c7cf507 into 3f9225d - view on LGTM.com

new alerts:

• 1 for FIXME comment

[lgtm-com]

This pull request introduces 1 alert when merging f8667a4 into e65acaa - view on LGTM.com

new alerts:

• 1 for FIXME comment

[jimklimov]

Some investigation showed that the token we use is defined in current net-snmp headers as a legacy pointer type, oid *, while many other names in the same header are arrays typed like oid[10]. There are also named macros to represent the amount of entries in those arrays, which is what we would use as a priority if available.

[Bi11]

It is more than a warning. Because incorrect length of usmAES128PrivProtocol, generate_Ku fails and AES doesn't work.
The similar issue is affecting nut-scanner too.

See #734 (comment)

[jimklimov]

Hell @Bi11 and sorry for the delay. I've read through this and linked comments, and my understanding is that with the current state of the master branch (or at least with these fixes applied to your build) AES works for you, while in the (aged) 2.7.4 release it did not?

Can you clarify please if there remains something to fix in nut-scanner along the same lines, or is it also already fixed? In the past weeks I juggled almost a thousand commits to fix warnings here and there, and lost track a bit :) So verifications like this are very welcome, whatever the verdict ;)

[Bi11]

I've just tried the master branch, the "sizeof" issue in 2.7.4 has been fixed with your commits, both in drivers/snmp-ups and nut-scanner. I'm really appreciated for your improvements made to this project, @jimklimov.

However, nut-scanner is still not working with my UPS, and there are some other trivial issues. I'll create a new Issue or PR once I make it work with my UPS.

[jimklimov]

Thank you very much for confirmations! I know about 1200 warnings remain per clang-9, though most are relatively trivial double-vs-float and int-unsigned-size_t which should not be fatal. A few left me scratching my head, as attempts to quickly fix grew into a big can of worms to pull. So I wanted to fix others first and then better see impact of remaining efforts, like when warnings appear in previously green code :) If you want to join this crusade, note the "fightwarn" branch often contains a few fixes ahead of those committed to master.

…

On Sat, Dec 12, 2020, 17:21 Bi11 ***@***.***> wrote: I've just tried the master branch, the "sizeof" issue in 2.7.4 has been fixed with your commits, both in drivers/snmp-ups and nut-scanner. I'm really appreciated for your improvements made to this project, @jimklimov <https://github.com/jimklimov>. However, nut-scanner is still not working with my UPS, and there are some other trivial issues. I'll create a new Issue or PR once I make it work with my UPS. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#850 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMPTFCUZ5IOWEXTXCFTABDSUOJ75ANCNFSM4TOEKDQA> .