Hi,

On 01/31/2017 07:46 PM, Adrian Bunk wrote:
> On Tue, Jan 31, 2017 at 04:02:53PM +0100, Philipp Kern wrote:
>> so tpm-utils 1.3.9 fixes OpenSSL 1.1 compatibility
> does 1.3.9 compile for you with OpenSSL 1.1?

1.3.9 still has two issues with -Werror, but none with OpenSSL 1.1 AFAICS.

Specifically it still needs this:

> Index: tpm-tools/src/tpm_mgmt/tpm_nvcommon.c
> ===================================================================
> --- tpm-tools.orig/src/tpm_mgmt/tpm_nvcommon.c
> +++ tpm-tools/src/tpm_mgmt/tpm_nvcommon.c
> @@ -164,8 +164,8 @@ int parseStringWithValues(const char *aA
>                               return -1;
>                       }
>  
> -                     if (!aArg[offset+numbytes] == '|' &&
> -                         !aArg[offset+numbytes] == 0) {
> +                     if (aArg[offset+numbytes] == '|' ||
> +                         aArg[offset+numbytes] == 0) {
>                               logError(_("Illegal character following decimal 
> "
>                                          "number in %s\n"),
>                                        aArg + offset);
> Index: tpm-tools/src/tpm_mgmt/tpm_present.c
> ===================================================================
> --- tpm-tools.orig/src/tpm_mgmt/tpm_present.c
> +++ tpm-tools/src/tpm_mgmt/tpm_present.c
> @@ -357,5 +357,5 @@ out_close:
>  out:
>      if (szTpmPasswd && !isWellKnown)
>       shredPasswd( szTpmPasswd );
> -     return iRc;
> +    return iRc;
>  }

I'd hold that -Werror is a problem in itself, but the comparison warning
was actually appreciated:

> tpm_nvcommon.c: In function ‘parseStringWithValues’:
> tpm_nvcommon.c:167:31: warning: comparison of constant ‘124’ with boolean 
> expression is always false [-Wbool-compare]
>     if (!aArg[offset+numbytes] == '|' &&
>                                ^~
> tpm_nvcommon.c:167:31: warning: logical not is only applied to the left hand 
> side of comparison [-Wlogical-not-parentheses]

So what happened is that the upstream maintainer only fixed one of the
two identical comparisons (the one for hexadecimals, not the one for
decimals).

With that I think we should be fine to upload this. It's quite late,
though, so it will need a little convincing of the Release Team.

Kind regards
Philipp Kern

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to