[rs-dev] Fwd: [rs-commit] r431 - /mod_timestamp/trunk/mod_timestamp.c

[Dirk-Willem van Gulik]

This broke builds on OSX. Is this the right strategy - or should I move to some other way of keeping lists/arrays of EVP_Digests ?

Am checking the other modules too - as this broke OSX ports for BigSur.

With kind regards,

Dw

> Begin forwarded message:
> 
> From: rs-commit--- via rs-commit <rs-commit at redwax.eu>
> Subject: [rs-commit] r431 - /mod_timestamp/trunk/mod_timestamp.c
> Date: 21 January 2022 at 21:53:50 CET
> To: rs-commit at redwax.eu
> Reply-To: rs-commit at redwax.eu
> 
> Author: dirkx at redwax.eu
> Date: Fri Jan 21 21:53:49 2022
> New Revision: 431
> 
> Log:
> DEFINE_STACK_OF(EVP_MD) seems to have gone; recreate it. And quell a warning.
> 
> Modified:
>    mod_timestamp/trunk/mod_timestamp.c
> 
> Modified: mod_timestamp/trunk/mod_timestamp.c
> ==============================================================================
> --- mod_timestamp/trunk/mod_timestamp.c	(original)
> +++ mod_timestamp/trunk/mod_timestamp.c	Fri Jan 21 21:53:49 2022
> @@ -40,6 +40,14 @@
> #include "util_script.h"
> 
> #include "mod_ca.h"
> +
> +#ifndef sk_EVP_MD_free
> +/* Recent versions of OpenSSL seem to no longer define 
> + * a stack of EVP_MD's. 
> + */
> +#include <openssl/safestack.h>
> +DEFINE_STACK_OF(EVP_MD)
> +#endif
> 
> #define DEFAULT_TIMESTAMP_SIZE 128*1024
> 
> @@ -360,7 +368,7 @@
>         return apr_psprintf(cmd->pool,
>                 "'%s' could not be recognised as a valid digest.", arg);
>     }
> -    if (!sk_EVP_MD_push(conf->digests, digest)) {
> +    if (!sk_EVP_MD_push(conf->digests, (EVP_MD *)digest)) {
>         return apr_psprintf(cmd->pool,
>                 "'%s' could not be added as a valid digest.", arg);
>     }
> 
> _______________________________________________
> rs-commit mailing list
> rs-commit at redwax.eu
> https://redwax.eu/mailman/listinfo/rs-commit
>