From: Bob Blick <bobblick@outlook.com>
To: Microcontroller discussion list - Public. <piclist@mit.edu>
Sender: "piclist-bounces@mit.edu" <piclist-bounces@mit.edu>
Date: Sat, 1 Aug 2020 18:25:30 -0700
Subject: Re: [PIC] Simple GCC question
Thread-Topic: [PIC] Simple GCC question
Thread-Index: AQHWZotc+nNhZovUu0iBK4Bhj1/IWakgpd6AgAAPvzOAAK98gIACo93A
Message-ID:
 <BYAPR02MB4055FF1BBEE71EFC97D6B1CADC4C0@BYAPR02MB4055.namprd02.prod.outlook.com>
References:
 <BYAPR02MB405521AE3E1393B371E463FDDC710@BYAPR02MB4055.namprd02.prod.outlook.com>,
	<f5e6b4c6-f8de-1312-e661-84ac5e2ff61d@gmail.com>
	<BYAPR02MB4055A035A18CB0FA06DD5677DC710@BYAPR02MB4055.namprd02.prod.outlook.com>,
	<alpine.BSF.2.21.9999.2007310927390.29099@nickel.allotrope.net>
In-Reply-To: <alpine.BSF.2.21.9999.2007310927390.29099@nickel.allotrope.net>
Reply-To: Microcontroller discussion list - Public. <piclist@mit.edu>
Accept-Language: en-US
Content-Language: en-US
X-MS-Exchange-Organization-SenderIdResult: TempError
X-MS-Exchange-Organization-PRD: mit.edu
X-MS-TNEF-Correlator:
acceptlanguage: en-US
received-spf: Pass (protection.outlook.com: domain of outlook.com designates
	40.92.4.68 as permitted sender) receiver=protection.outlook.com;
	client-ip=40.92.4.68; helo=NAM02-CY1-obe.outbound.protection.outlook.com;
dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=outlook.com;
	s=selector1;
	h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck;
	bh=Rw/RbBIBBwK6xdzpKDWjonDgO9wtxS9CAlx9ISWiYp4=;
	b=FSyBbvpXn8hQ4CPyJOZM7t/tivUmRscTVP9YBvghxWwYZCKPsteD9e9I0nkOc09m0ImXXLn20NaWSNAOnlP1ebZqgyHv7SLE5Jz9rGC7Tu8iMHRwhN3SKoH+Hs+VnqmUEc9fgJG0iunsidLg/SJqZHhj6c3s65ajnTFZVqUR9/8TQXBvs4Vor/6eDHEpEfZOxS2ljgc8iFAoKhW0eI4HYXeL9Oqzz+rkPdJ6TrstcY7tX+qsa+VEGN3zlQHup9/xocgNbaV0mfTL/XqY22f8oRpqb24j2QuCPqEotBp0bORVmmix0JbRfoUiNSeA8GPn94NkFBUhEGb1nh28Rbk+Fg==
authentication-results: spf=pass (sender IP is 40.92.4.68)
	smtp.mailfrom=outlook.com; mit.edu; dkim=pass (signature was verified)
	header.d=outlook.com; mit.edu;	dmarc=pass action=none
 header.from=outlook.com; 
errors-to: piclist-bounces@mit.edu
list-id: "Microcontroller discussion list - Public." <piclist.mit.edu>
list-post: <mailto:piclist@mit.edu>
x-beenthere: piclist@mit.edu
x-mailman-version: 2.1.6
x-topics: [PIC]
x-mime-autoconverted: from quoted-printable to 8bit by PCH.mit.edu id
	0721PZpo017378
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Hi Sergio,

Thank you very much for this. I appreciate the detail, since I am not a nat=
ural programmer, it helps me a lot to see the right way to move when I impr=
ove this library. Which I will when I do the next project that needs it. Ri=
ght now I must press on to finish this one.
Many thanks.

Bob

________________________________________
From: piclist-bounces@mit.edu <piclist-bounces@mit.edu> on behalf of sergio=
=20
Sent: Friday, July 31, 2020 2:00 AM
To: Microcontroller discussion list - Public.
Subject: Re: [PIC] Simple GCC question

Sorry I'm late to the party but for future reference this might help:

Try to convert the offending function to a macro and use that to
initialise your structure.

Looking at the "Lcd_create" function I can see that all it is doing is
initialising a copy of the structure, calling "Lcd_init" with a
pointer to that copy and then returning the copy to be copied into the
actual "lcd" structure (all (mostly) incredibly redundent and inefficent).

What I would do in this instance is remove the call to "Lcd_init" from
"Lcd_create" and then convert "Lcd_create" to a macro thus:

#define LCD_CREATE(                     \
                port, pin,              \
                rs_port, rs_pin,        \
                en_port, en_pin, mode)  \
        {                               \
                .mode =3D mode,           \
                                        \
                .en_pin =3D en_pin,       \
                .en_port =3D en_port,     \
                                        \
                .rs_pin =3D rs_pin,       \
                .rs_port =3D rs_port,     \
                                        \
                .data_pin =3D pin,        \
                .data_port =3D port       \
        }

BEWARE: the "\" at the end of the line should be the last character
before the newline. An unseen space such as "\ " will cause problems

Your code should then change to this:

Lcd_PortType
        ports[] =3D
        {       LCD_D4_GPIO_Port,
                LCD_D5_GPIO_Port,
                LCD_D6_GPIO_Port,
                LCD_D7_GPIO_Port
        };

Lcd_PinType
        pins[] =3D
        {       LCD_D4_Pin,
                LCD_D5_Pin,
                LCD_D6_Pin,
                LCD_D7_Pin
        };

/* can't remember if you need the "\" below */
Lcd_HandleTypeDef
        lcd =3D LCD_CREATE(ports,                 \
                         pins,                  \
                         LCD_RS_GPIO_Port,      \
                         LCD_RS_Pin,            \
                         LCD_EN_GPIO_Port,      \
                         LCD_EN_Pin,            \
                         LCD_4_BIT_MODE);

void main(void)
{
        /* Initialise the lcd here */
        Lcd_init(&lcd);

        /* Other stuff */
        for(;;);
}

I haven't actually tried compiling this (too much effort) but it should
work :-)

Regards
Sergio Masci


On Fri, 31 Jul 2020, Bob Blick wrote:

> Hi Mike,
>
> That was simple, and not only did it compile, but my LCD display still wo=
rks.
>
> Thank you so much and thanks for all the other answers!
>
> Bob
>
> ________________________________________
> From: piclist-bounces@mit.edu <piclist-bounces@mit.edu> on behalf of Mike
> Sent: Thursday, July 30, 2020 2:36 PM
> To: piclist@mit.edu
> Subject: Re: [PIC] Simple GCC question
>
> I strongly suspect the non-constant value GCC is complaining about is
> the return value from Lcd_create().  Since the lcd structure is being
> initialased via a function call you cannot do this outside of a
> function.  You can however define an instance of the lcd struct outside
> of a function, and then initialise it later.
>
> Something like this should work:
>
> main.c
>
> #include <lcd.h>
>
> /* These can be initialised outside a function since the initialisers are=
 constant */
> Lcd_PortType ports[] =3D {LCD_D4_GPIO_Port, LCD_D5_GPIO_Port, LCD_D6_GPIO=
_Port, LCD_D7_GPIO_Port};
> Lcd_PinType pins[] =3D {LCD_D4_Pin, LCD_D5_Pin, LCD_D6_Pin, LCD_D7_Pin};
> /* You can only define an instance of Lcd_HandleTypeDef here, since the i=
nitialiser is a function */
> Lcd_HandleTypeDef lcd;
>
> void main(void)
> {
>   /* Initialise the lcd here */
>   lcd =3D Lcd_create(ports, pins, LCD_RS_GPIO_Port, LCD_RS_Pin, LCD_EN_GP=
IO_Port, LCD_EN_Pin, LCD_4_BIT_MODE);
>
>   /* Other stuff */
>   for(;;);
> }
>
>
> mymodule.c
>
> #include <lcd.h>
>
> /* Declare the lcd struct defined in main  */
> extern Lcd_HandleTypeDef lcd;
>
> void myfunction(void)
> {
>   Lcd_clear(&lcd);
> }
>
>
> This is the simplest way I could show this, but using global variables
> isn't typically the nicest way to do things.
>
> Regards
>
> Mike
>
>
> On 30/07/2020 17:26, Bob Blick wrote:
>> I'm using a nice little HD44780 LCD library in an STM32 project, it work=
s great but I'm stuck with a variable scope problem that I can only seem to=
 make worse because GCC is a bit stricter than I want it to be at this poin=
t. Sorry if the code example doesn't wordwrap properly.
>>
>> If, somewhere in main() I create an instance of the LCD like this:
>>    Lcd_PortType ports[] =3D {LCD_D4_GPIO_Port, LCD_D5_GPIO_Port, LCD_D6_=
GPIO_Port, LCD_D7_GPIO_Port};
>>    Lcd_PinType pins[] =3D {LCD_D4_Pin, LCD_D5_Pin, LCD_D6_Pin, LCD_D7_Pi=
n};
>>    Lcd_HandleTypeDef lcd =3D Lcd_create(ports, pins, LCD_RS_GPIO_Port, L=
CD_RS_Pin, LCD_EN_GPIO_Port, LCD_EN_Pin, LCD_4_BIT_MODE);
>>
>> then I can use it like this:
>>    Lcd_cursor(&lcd, 1,6);
>>
>> but of course I can't do that from within another function since "lcd" o=
nly exists in main()
>>
>> If I try to make it all global by creating lcd outside of main(), GCC do=
esn't like that because initializer elements must be constants and it doesn=
't consider "ports" and "pins" to be constants.
>>
>> Not being a real C programmer this has stopped me in my tracks. If I was=
 stuck on a deserted island I would use this library as a basis for somethi=
ng with ports and pins hardcoded, but I'm hoping a real expert would give m=
e some gentle help, which will certainly educate me about doing things the =
correct way and also get me past this little hurdle.
>>
>> Thanks, Bob
>


--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.