Re: [openssl-project] OS/X builds failing

[Viktor Dukhovni]

> On Feb 9, 2018, at 12:39 PM, Kurt Roeckx  wrote:
> 
> We document printf to be like the C standard, so it should be
> intmax_t, not int64_t, or we need to fix the documentation.

I don't think we get to coöpt "j" for another purpose, if we
we want a letter format for int64t, we'll need to use a new
one, but that's a bad idea.  So I think the "or ..." part is
not a good idea.

Just cast to (intmax_t) or use OSSL_PRIi64/OSSL_PRIu64 (defined
to PRIi64/PRIu64 where available or else to some appropriate
letter for the platform in question).

BIO_printf should handle the underlying primitive types when,
e.g., encountering "%lu", "%llu", "%ju" and "%zu"...

-- 
Viktor.

___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Re: [openssl-project] OS/X builds failing

[Viktor Dukhovni]

> On Feb 9, 2018, at 11:23 AM, Richard Levitte  wrote:
> 
> From those errors, it looks to me like uintmax_t isn't 64-bit on that
> Mac OS/X machine, unless 'unsigned long' and 'unsigned long long' are
> the same.

No, the compiler is not telling you they're not actually the same, rather
it is telling you that their definitions are independent, and they *could*
be different, and so the code is not portable.

> (that error does surprise me, since uintmax_t is supposed to be the
> largest integer type possible, and yet, the error suggests that this
> isn't the case here)

See above.  Compiling the below on the latest MacOS/X:

$ cat size.c
#include 
#include 

int main(int argc, char **argv)
{
printf("char = %zu\n", sizeof(unsigned char));
printf("short = %zu\n", sizeof(unsigned short));
printf("int = %zu\n", sizeof(unsigned int));
printf("long = %zu\n", sizeof(unsigned long int));
printf("long long = %zu\n", sizeof(unsigned long long int));

printf("int8_t = %zu\n", sizeof(uint8_t));
printf("int16_t = %zu\n", sizeof(uint16_t));
printf("int32_t = %zu\n", sizeof(uint32_t));
printf("int64_t = %zu\n", sizeof(uint64_t));
printf("intmax_t = %zu\n", sizeof(uintmax_t));
return 0;
}

shows that long, long long, int64_t and intmax_t are presently all the same:

$ ./size
char = 1
short = 2
int = 4
long = 8
long long = 8
int8_t = 1
int16_t = 2
int32_t = 4
int64_t = 8
intmax_t = 8

-- 
Viktor.

___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Re: [openssl-project] OS/X builds failing

[Viktor Dukhovni]

> On Feb 9, 2018, at 5:15 AM, Matt Caswell  wrote:
> 
> The new travis OS/X builds are failing with this:
> 
> -MT apps/enc.o -c -o apps/enc.o apps/enc.c
> apps/enc.c:567:54: error: format specifies type 'uintmax_t' (aka
> 'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
> long long') [-Werror,-Wformat]
>BIO_printf(bio_err, "bytes read   : %8ju\n", BIO_number_read(in));
> ^~~
>%8llu
> apps/enc.c:568:54: error: format specifies type 'uintmax_t' (aka
> 'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
> long long') [-Werror,-Wformat]
>BIO_printf(bio_err, "bytes written: %8ju\n",
> BIO_number_written(out));
> ^~~
>%8llu
> 2 errors generated.
> 
> 
> The thing is though this is *our* BIO_printf, and we do seem to expect a
> uint64_t for a "%ju". So I'm not sure how to fix that.

The obvious and correct fix is to cast the output of BIO_number_read() and
BIO_number_written() to uintmax_t.  And we should expect uintmax_t with %j,
even in our BIO_printf().  It happens to be 64-bit on all the platforms we
support perhaps (any Crays?), but that's just a coincidence.

-- 
Viktor.

___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Re: [openssl-project] OS/X builds failing

[Andy Polyakov]

For the record. I don't generally appreciate fast commits, but I feel
like 100 times stronger when it comes to public headers [naturally in
supported or minor release] and I can't qualify 19 minutes turnaround
for merge request as appropriate.
___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Re: [openssl-project] OS/X builds failing

[Andy Polyakov]

>> apps/enc.c:568:54: error: format specifies type 'uintmax_t' (aka
>> 'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
>> long long') [-Werror,-Wformat]
>> BIO_printf(bio_err, "bytes written: %8ju\n",
>> BIO_number_written(out));
>>  ^~~
>> %8llu
>> 2 errors generated.
>>
>>
>> The thing is though this is *our* BIO_printf, and we do seem to expect a
>> uint64_t for a "%ju". So I'm not sure how to fix that.
> 
> I'd suggest -Wno-format. With rationale that a) Xcode is arguably
> somewhat unreasonable, b) we can rely on Linux builds to catch warnings
> that actually matter.

Another possibility *could* be suppressing corresponding __attribute__
in bio.h, but it's public header, so question would be if change can be
justified in minor release. Just in case, suppress specifically for
Xcode [which would cover even iOS].
___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

Re: [openssl-project] OS/X builds failing

[Andy Polyakov]

> apps/enc.c:568:54: error: format specifies type 'uintmax_t' (aka
> 'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
> long long') [-Werror,-Wformat]
> BIO_printf(bio_err, "bytes written: %8ju\n",
> BIO_number_written(out));
>  ^~~
> %8llu
> 2 errors generated.
> 
> 
> The thing is though this is *our* BIO_printf, and we do seem to expect a
> uint64_t for a "%ju". So I'm not sure how to fix that.

I'd suggest -Wno-format. With rationale that a) Xcode is arguably
somewhat unreasonable, b) we can rely on Linux builds to catch warnings
that actually matter.
___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project

[openssl-project] OS/X builds failing

[Matt Caswell]

The new travis OS/X builds are failing with this:

-MT apps/enc.o -c -o apps/enc.o apps/enc.c
apps/enc.c:567:54: error: format specifies type 'uintmax_t' (aka
'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
long long') [-Werror,-Wformat]
BIO_printf(bio_err, "bytes read   : %8ju\n", BIO_number_read(in));
 ^~~
%8llu
apps/enc.c:568:54: error: format specifies type 'uintmax_t' (aka
'unsigned long') but the argument has type 'uint64_t' (aka 'unsigned
long long') [-Werror,-Wformat]
BIO_printf(bio_err, "bytes written: %8ju\n",
BIO_number_written(out));
 ^~~
%8llu
2 errors generated.


The thing is though this is *our* BIO_printf, and we do seem to expect a
uint64_t for a "%ju". So I'm not sure how to fix that.

Matt

___
openssl-project mailing list
openssl-project@openssl.org
https://mta.openssl.org/mailman/listinfo/openssl-project