Skip to contentSkip to conversation

checkpatch suspected false positive

DISCUSSION Tobin C. Harding 2 participants 3 messages Feb 21, 2017

lore
Share & tools
lore
#0 lore raw

Checkpatch may be giving a false positive of type CONST_STRUCT when parsing files in drivers/staging/comedi/drivers.

$ pwd
build/kernel/linux-trees/gregKH/staging/
$ cd drivers/staging/comedi/drivers
$ checkpatch --terse --show-types *.c | grep CONST_STRUCT
addi_apci_3501.c:97: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:972: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:1006: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:659: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:667: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:668: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
ni_670x.c:212: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const

snippet from das16.c

/* get any user-defined input range */

if (pg_type == das16_pg_none && (min || max)) {
	struct comedi_lrange *lrange;
	struct comedi_krange *krange;

	/* allocate single-range range table */
	lrange = comedi_alloc_spriv(s, sizeof(*lrange) + sizeof(*krange));
	if (!lrange)
		return &range_unknown;

	/* initialize ai range */
	lrange->length = 1;
	krange = lrange->range;
	krange->min = min;
	krange->max = max;
	krange->flags = UNIT_volt;

	return lrange;
}
Quoted 1 line

From snippet it may be seen that struct comedi_lrange *lrange should

not be const.

In the event that I am in the wrong and checkpatch is correct please feel free to bluntly correct me.

thanks,
Tobin.

#1 lore raw
Quoted 8 lines · Tobin C. Harding

Checkpatch may be giving a false positive of type CONST_STRUCT when parsing files in drivers/staging/comedi/drivers.

$ pwd
build/kernel/linux-trees/gregKH/staging/
$ cd drivers/staging/comedi/drivers
Quoted 8 lines
$ checkpatch --terse --show-types *.c | grep CONST_STRUCT
addi_apci_3501.c:97: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:972: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:1006: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:659: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:667: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:668: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
ni_670x.c:212: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const

checkpatch is brainless, it just looks for patterns that are atypical.

$ git grep -E "struct\s+comedi_lrange\b" | wc -l
223
$ git grep -E "const\s+struct\s+comedi_lrange\b" | wc -l
215

So, yes, that struct is normally const.
Normally doesn't mean always or has to be.

#2 lore raw
Quoted 21 lines · Joe Perches
Quoted 8 lines · Tobin C. Harding

Checkpatch may be giving a false positive of type CONST_STRUCT when parsing files in drivers/staging/comedi/drivers.

$ pwd
build/kernel/linux-trees/gregKH/staging/
$ cd drivers/staging/comedi/drivers
Quoted 8 lines
$ checkpatch --terse --show-types *.c | grep CONST_STRUCT
addi_apci_3501.c:97: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:972: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
das16.c:1006: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:659: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:667: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
jr3_pci.c:668: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const
ni_670x.c:212: WARNING:CONST_STRUCT: struct comedi_lrange should normally be const

checkpatch is brainless, it just looks for patterns that are atypical.

Quoted 8 lines
$ git grep -E "struct\s+comedi_lrange\b" | wc -l
223
$ git grep -E "const\s+struct\s+comedi_lrange\b" | wc -l
215

So, yes, that struct is normally const.
Normally doesn't mean always or has to be.

Cheers Joe. I'm sure this is not the first time you have explained that. Would it be a good idea to add some documentation (in Documentation/process) about checkpatch gotchas. Perhaps we could save some people some time if every newbie didn't have to make the same mistakes (or ask the same questions).

The first two could be;

1. Don't fix line over 80 warnings if it does not objectively make the code more readable (see coding-style.rst, grep for 'Statements longer than 80 columns').

2. Warning struct foo should normally be const ... description of how checkpatch decides this warning is necessary.

thanks,
Tobin.