topic Re: CodeWarrior for HC12 version 5.2 does not comply to C90 in CodeWarrior for MCU

CodeWarrior for HC12 version 5.2 does not comply to C90

robertobagnara — Tue, 16 Apr 2019 12:34:29 GMT

Compiling the following snippet, say, included in source file r.c:

int a[sizeof(char) == 2 ? -1 : 1];

with a command like

"C:\Program Files (x86)\Freescale\CWS12v5.2\Prog\chc12.exe" -Tc2s2 r.c

shows that the compiler violates C90 as far as the very important clause 6.3.3.4 is concerned, where it says:

When applied to an operand that has type char. unsigned char. or signed char. (or a qualified version thereof) the result is 1.

Is this a known bug?

Re: CodeWarrior for HC12 version 5.2 does not comply to C90

ZhangJennie — Wed, 17 Apr 2019 09:47:20 GMT

Hi Roberto Bagnara

I tested your code on my side, there is neither build error nor debug error. see attached video.

please create a new project with IDE and test it also.

Have a great day,
Jun Zhang

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------

Re: CodeWarrior for HC12 version 5.2 does not comply to C90

robertobagnara — Wed, 17 Apr 2019 18:01:45 GMT

Hi Jun Zhang.

You probably missed the -Tc2s2 option on the command line. This sets the size of both chars and shorts to 16 bits. The option is described in the manual in section "-T: Flexible Type Management". Figure 5.4 in the same manual explains how to set this in the IDE. The bug in the compiler manifests itself only if the size of chars is set to anything different from 8 bits.

Kind regards,

Roberto

Re: CodeWarrior for HC12 version 5.2 does not comply to C90

kef2 — Sat, 20 Apr 2019 18:15:09 GMT

shows that the compiler violates C90 as far as the very important clause 6.3.3.4 is concerned, where it says:

When applied to an operand that has type char. unsigned char. or signed char. (or a qualified version thereof) the result is 1.

It also says

The sizeof operator yields the size (in bytes) of its operand, which may be an expression or the parenthesized name of a type.

Bytes, not chars. And both clauses about sizeof kind of forbid option to make char wider than minimum addressable piece of memory (8bits byte on S12(X), 16bits byte on TI's C2000).