Skip to contentSkip to conversation

Staging: comedi: serial2002: fixed consistent spacing issue

PATCH In review Alice Ferrazzi 5 participants 8 messages Mar 20, 2013

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Staging: comedi: serial2002: fixed consistent spacing issue

Fixed consistent spacing around '*'.

---
 drivers/staging/comedi/drivers/serial2002.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/comedi/drivers/serial2002.c +1 -1
index e6177b4..858d364 100644
@@ -604,7 +604,7 @@ static int serial_2002_open(struct comedi_device *dev)
 							    c[j].max;
 							range_table_list[chan] =
 							    (const struct
-							     comedi_lrange *)
+							     comedi_lrange*)
 							    &range[j];
 						}
 						maxdata_list[chan] =
Signature · 1 line
1.7.10.4
#1 lore raw
Quoted 2 lines · Alice Ferrazzi

Fixed consistent spacing around '*'.

The original was correct, actually.

Quoted 10 lines
---
 drivers/staging/comedi/drivers/serial2002.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/comedi/drivers/serial2002.c +0 -0
index e6177b4..858d364 100644

drivers/staging/comedi/drivers/serial2002.c static int serial_2002_open(struct comedi_device *dev) · line 608

Quoted 6 lines
drivers/staging/comedi/drivers/serial2002.c +1 -1
@@ -604,7 +604,7 @@ static int serial_2002_open(struct comedi_device *dev)
 							    c[j].max;
 							range_table_list[chan] =
 							    (const struct
-							     comedi_lrange *)
+							     comedi_lrange*)
 							    &range[j];

The original code here needs to broken up into functions so it isn't squashed up against the 80 character limit.

For casts the spacing should look like:

	foo = (struct my_struct *)ptr;

There is a space after "my_struct" but no space after the closing parenthesis. Use that to remind yourself that casting is a high precedence operation.

For declaring pointers the spacing is:

	struct my_struct *ptr;

For multiplication the spacing is:

	foo = x * y;

Or multplication with a dereference it would be:

	foo = x * *ptr;

regards,
dan carpenter

#2 lore raw
Quoted 2 lines · Dan Carpenter

The original code here needs to broken up into functions so it isn't squashed up against the 80 character limit.

I'd say what needs to be done to the original code...

Observation 1:
if (foo) {
A /* two lines */
} else {
B /* huge pile of shite */
}
return result;

is equivalent to

if (foo) {
	A
	return result;
}

B
return result;

Observation 2:

while (1) {
	A	/* a couple of lines */
	if (foo) {
		break;
	} else {
		B /* huge pile of shite */
	}
}
is equivalent to
while (1) {
	A
	if (foo)
		break;
	B
}

Observation 3:

while (1) {
	A	/* moderate pile of shite, assigning foo */
	if (foo) {
		B /* huge pile of shite */
	}
}
is equivalent to
while (1) {
	A
	if (!foo)
		continue;
	B
}

Observation 4: functions are there for purpose. When you have two identical piles of garbage (avert your eyes, or risk taking another look at your dinner) such as

				int unit, sign, min;
				unit =
				    (data.value >> 10) &
				    0x7;
				sign =
				    (data.value >> 13) &
				    0x1;
				min =
				    (data.value >> 14) &
				    0xfffff;

				switch (unit) {
				case 0:{
						min =
						    min
						    *
						    1000000;
					}
					break;
				case 1:{
						min =
						    min
						    *
						    1000;
					}
					break;
				case 2:{
						min =
						    min
						    * 1;
					}
					break;
				}
				if (sign)
					min = -min;

you just might consider turning that pile of excrements into a helper function. Incidentally, min = min * 1 is somewhat, er, pointless...

Observation 5:
for (i = 0; i <= 4; i++) {

{
	switch (i) {
	case 0: c = non_NULL_1; ... break;
	case 1: c = non_NULL_2; ... break;
	case 2: c = non_NULL_3; ... break;
	case 3: c = non_NULL_4; ... break;
	case 4: c = non_NULL_5; ... break;
	default: c = NULL; break;
	}
	if (c) {
		pile_of_shite
	}
}

might, perhaps, be taking defensive programming a bit too far...

Observation 6:
the Vogon whose brain has produced that code up had been brought up on Pascal, Ada or something worse, and had been badly traumatized by semantics of switch and break.

switch (foo) {
	case 0: {
		bar = baz;
	} break;
	case 1: {
	.....
}

is not quite conventional for C.

Observation 7: down, not across...

#3 lore raw
Quoted 6 lines · Al Viro
Quoted 2 lines · Dan Carpenter

The original code here needs to broken up into functions so it isn't squashed up against the 80 character limit.

I'd say what needs to be done to the original code...

All good things, thanks for taking the time to write it Al. Not that I'm ever touching comedi code, but Hartley is, so I'm adding him to cc just in case he didn't see it...

Quoted 122 lines

Observation 1:
if (foo) {
A /* two lines */
} else {
B /* huge pile of shite */
}
return result;

is equivalent to

if (foo) {
	A
	return result;
}

B
return result;

Observation 2:

while (1) {
	A	/* a couple of lines */
	if (foo) {
		break;
	} else {
		B /* huge pile of shite */
	}
}
is equivalent to
while (1) {
	A
	if (foo)
		break;
	B
}

Observation 3:

while (1) {
	A	/* moderate pile of shite, assigning foo */
	if (foo) {
		B /* huge pile of shite */
	}
}
is equivalent to
while (1) {
	A
	if (!foo)
		continue;
	B
}

Observation 4: functions are there for purpose. When you have two identical piles of garbage (avert your eyes, or risk taking another look at your dinner) such as

				int unit, sign, min;
				unit =
				    (data.value >> 10) &
				    0x7;
				sign =
				    (data.value >> 13) &
				    0x1;
				min =
				    (data.value >> 14) &
				    0xfffff;

				switch (unit) {
				case 0:{
						min =
						    min
						    *
						    1000000;
					}
					break;
				case 1:{
						min =
						    min
						    *
						    1000;
					}
					break;
				case 2:{
						min =
						    min
						    * 1;
					}
					break;
				}
				if (sign)
					min = -min;

you just might consider turning that pile of excrements into a helper function. Incidentally, min = min * 1 is somewhat, er, pointless...

Observation 5:
for (i = 0; i <= 4; i++) {

{
	switch (i) {
	case 0: c = non_NULL_1; ... break;
	case 1: c = non_NULL_2; ... break;
	case 2: c = non_NULL_3; ... break;
	case 3: c = non_NULL_4; ... break;
	case 4: c = non_NULL_5; ... break;
	default: c = NULL; break;
	}
	if (c) {
		pile_of_shite
	}
}

might, perhaps, be taking defensive programming a bit too far...

Observation 6:
the Vogon whose brain has produced that code up had been brought up on Pascal, Ada or something worse, and had been badly traumatized by semantics of switch and break.

switch (foo) {
	case 0: {
		bar = baz;
	} break;
	case 1: {
	.....
}

is not quite conventional for C.

Observation 7: down, not across...

#4 lore raw
Quoted 11 lines · Joe Perches
Quoted 6 lines · Al Viro
Quoted 2 lines · Dan Carpenter

The original code here needs to broken up into functions so it isn't squashed up against the 80 character limit.

I'd say what needs to be done to the original code...

All good things, thanks for taking the time to write it Al. Not that I'm ever touching comedi code, but Hartley is, so I'm adding him to cc just in case he didn't see it...

I saw it.

I haven't touched this driver yet because I haven't worked out a clean way to fix the sparse warnings in it.

I'll get to it eventually. :-)

Regards,
Hartley

#5 lore raw

PATCH 1/1 CodingStyle: Add tab indentation avoidance tips

In reply to Al Viro

Add Al's comments in from https://lkml.org/lkml/2013/3/20/345

---

Documentation/CodingStyle +135 -0
index e00b8f0..c4ba183 100644
@@ -836,6 +836,141 @@ next instruction in the assembly output:
 	     : /* outputs */ : /* inputs */ : /* clobbers */);
 
 
+		Chapter 20:
+
+Tips to avoid overly tab indented code.
+
+Tip 1:
+	if (foo) {
+		A /* two lines */
+	} else {
+		B /* huge pile of shite */
+	}
+	return result;
+
+is equivalent to
+
+	if (foo) {
+		A
+		return result;
+	}
+	B
+	return result;
+
+Tip 2:
+	while (1) {
+		A	/* a couple of lines */
+		if (foo) {
+			break;
+		} else {
+			B /* huge pile of shite */
+		}
+	}
+
+is equivalent to
+
+	while (1) {
+		A
+		if (foo)
+			break;
+		B
+	}
+
+Tip 3:
+	while (1) {
+		A       /* moderate pile of shite, assigning foo */
+		if (foo) {
+			B /* huge pile of shite */
+		}
+	}
+
+is equivalent to
+
+	while (1) {
+		A
+		if (!foo)
+			continue;
+		B
+	}
+
+Tip 4:
+
+functions are there for purpose.  When you have two identical piles of
+garbage (avert your eyes, or risk taking another look at your dinner)
+such as
+				int unit, sign, min;
+				unit =
+				    (data.value >> 10) &
+				    0x7;
+				sign =
+				    (data.value >> 13) &
+				    0x1;
+				min =
+				    (data.value >> 14) &
+				    0xfffff;
+
+				switch (unit) {
+				case 0:{
+						min =
+						    min
+						    *
+						    1000000;
+					}
+					break;
+				case 1:{
+						min =
+						    min
+						    *
+						    1000;
+					}
+					break;
+				case 2:{
+						min =
+						    min
+						    * 1;
+					}
+					break;
+				}
+				if (sign)
+					min = -min;
+
+you just might consider turning that pile of excrements into a helper
+function.  Incidentally, min = min * 1 is somewhat, er, pointless...
+
+Tip 5:
+	for (i = 0; i <= 4; i++) {
+	{
+		switch (i) {
+		case 0: c = non_NULL_1; ... break;
+		case 1: c = non_NULL_2; ... break;
+		case 2: c = non_NULL_3; ... break;
+		case 3: c = non_NULL_4; ... break;
+		case 4: c = non_NULL_5; ... break;
+		default: c = NULL; break;
+		}
+		if (c) {
+			pile_of_shite
+		}
+	}
+might, perhaps, be taking defensive programming a bit too far...
+
+Tip 6:
+The Vogon whose brain has produced that code up had been brought up on Pascal,
+Ada or something worse, and had been badly traumatized by semantics of switch
+and break.
+	switch (foo) {
+		case 0: {
+			bar = baz;
+		} break;
+		case 1: {
+		.....
+	}
+	is not quite conventional for C.
+
+Tip 7:
+Code flow is down, not across...
+
+
 
 		Appendix I: References
 
#6 lore raw

Re: [PATCH] CodingStyle: Add tab indentation avoidance tips

Documentation/CodingStyle next instruction in the assembly output: · line 971

Quoted 4 lines · Joe Perches

Add Al's comments in from https://lkml.org/lkml/2013/3/20/345

Documentation/CodingStyle +2 -0
+Tip 7:
+Code flow is down, not across...

That one is a lovely improvement on what I actually said. google "down, not across" for context...

#7 lore raw

Re: [PATCH] CodingStyle: Add tab indentation avoidance tips

Quoted 8 lines · Al Viro

Documentation/CodingStyle next instruction in the assembly output: · line 971

Quoted 4 lines · Joe Perches

Add Al's comments in from https://lkml.org/lkml/2013/3/20/345

Documentation/CodingStyle +2 -0
+Tip 7:
+Code flow is down, not across...

That one is a lovely improvement on what I actually said. google "down, not across" for context...

Heh. Context is everything.
I shoulda known better, given the source ;)

cheers, Joe