aboutsummaryrefslogtreecommitdiff
path: root/scripts/checkpatch.pl
AgeCommit message (Collapse)AuthorFilesLines
2014-06-04scripts/checkpatch.pl: device_initcall is not the only __initcall substituteFabian Frederick1-2/+2
This patch adds a link to init.h to find appropriate initcall function to replace obsolete __initcall Signed-off-by: Fabian Frederick <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: check stable email addressJoe Perches1-0/+6
It should be [email protected], not [email protected]. Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: warn on unnecessary void function return statementsJoe Perches1-0/+7
void function lines that use a single tab then "return;" are generally unnecessary. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: prefer kstrto<foo> to sscanf(buf, "%<lhuidx>", &bar);Joe Perches1-0/+21
Use the kstrto<foo> functions in preference to sscanf. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: add warning for kmalloc/kzalloc with multiplyJoe Perches1-0/+24
Protect against sizeof overflows by preferring kmalloc_array/kcalloc over kmalloc/kzalloc with a sizeof multiply. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: warn on #defines ending in semicolonJoe Perches1-0/+11
Using a #define ending in a semicolon is poor style and can lead to unexpected code paths being executed. Warn on uses of these #define types: #define foo[(...)] bar; #define foo[(...)] \ bar; Based on a patch from Borislav Petkov. Signed-off-by: Joe Perches <[email protected]> Cc: Borislav Petkov <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: make --strict a default for files in drivers/net and net/Joe Perches1-0/+13
Networking files are generally more strictly conformant to linux-kernel style so make checkpatch more verbose by default for patches to files or when checking files in these directories. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: David Miller <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: always warn on missing blank line after variable declaration blockJoe Perches1-10/+34
Make the test system wide, modify the message too. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-06-04checkpatch: fix wildcard DT compatible string checkingRob Herring1-2/+4
We attempt to search for compatible strings which use a variable token in the documented name such as <chip> or <soc>. While this was attempted to be handled, it's utterly broken. The desired forms of matching are: vendor,<chip>-* vendor,name<part#>-* For <chip>, lower case characters and numbers are permitted. For <part#>, only numeric values are allowed. With this change, the number of missing compatible strings reported in arch/arm/boot/dts is reduced from 1071 to 960. Reported-by: Alexandre Belloni <[email protected]> Signed-off-by: Rob Herring <[email protected]> Tested-by: Geert Uytterhoeven <[email protected]> Cc: Florian Vaussard <[email protected]> Cc: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: don't warn on bitfield spaces around :Joe Perches1-2/+5
This test prevents code from being aligned around the : for easy visual counting of bitfield lengths. ie: int foo : 1, int bar : 2, int foobar :29; should be acceptable so remove the test. Signed-off-by: Joe Perches <[email protected]> Suggested-by: Dan Carpenter <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: expand parenthesis alignment test to declarations, functions and ↵Joe Perches1-4/+6
assignments Currently the parenthesis alignment test works only on misalignments of if statements like if (foo(bar, baz) Expand the test to find misalignments like: static inline int foo(int bar, int baz) and foo(bar, baz); and foo = bar(baz, qux); Expand the $Inline keyword for __inline and __inline__ too. Add $Inline to $Declare so it also matches "static inline <foo>". These checks are only performed with --strict. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch.pl: add check for Change-IdChristopher Covington1-0/+6
A commit hook for the Gerrit code review server [1] inserts change identifiers so Gerrit can track patches through multiple revisions. These identifiers are noise in the context of the upstream kernel. (Many Gerrit servers are private. Even given a public instance, given only a Change-Id, one must guess which server a change was tracked on. Patches submitted to the Linux kernel mailing lists should be able to stand on their own. If it's truly useful to reference code review on a Gerrit server, a URL is a much clearer way to do so.) Thus, issue an error when a Change-Id line is encountered before the Signed-off-by. 1. https://gerrit.googlesource.com/gerrit/+/master/gerrit-server/src/main/resources/com/google/gerrit/server/tools/root/hooks/commit-msg Signed-off-by: Christopher Covington <[email protected]> Cc: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03scripts/checkpatch.pl: __GFP_NOFAIL isn't going awayAndrew Morton1-6/+0
Revert commit 7e4915e78992 ("checkpatch: add warning of future __GFP_NOFAIL use"). There are no plans to remove __GFP_NOFAIL. __GFP_NOFAIL exists to a) centralise the retry-allocation-for-ever operation into the core allocator, which is the appropriate implementation site and b) permit us to identify code sites which aren't handling memory exhaustion appropriately. Cc: David Rientjes <[email protected]> Cc: Joe Perches <[email protected]> Cc: Theodore Ts'o <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: net and drivers/net: warn on missing blank line after variable ↵Joe Perches1-0/+15
declaration Networking prefers this style, so warn when it's not used. Networking uses: void foo(int bar) { int baz; code... } not void foo(int bar) { int baz; code... } There are a limited number of false positives when using macros to declare variables like: WARNING: networking uses a blank line after declarations #330: FILE: net/ipv4/inet_hashtables.c:330: + int dif = sk->sk_bound_dev_if; + INET_ADDR_COOKIE(acookie, saddr, daddr) Signed-off-by: Joe Perches <[email protected]> Cc: David Miller <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: improve the compatible vendor matchFlorian Vaussard1-5/+5
Improve the vendor name match in vendor-prefix.txt by only matching the exact vendor name at the beginning of lines. Signed-off-by: Florian Vaussard <[email protected]> Cc: Joe Perches <[email protected]> Acked-by: Rob Herring <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: check compatible strings in .c and .h tooFlorian Vaussard1-2/+4
Look for ".compatible = "foo" strings not only in .dts files, but in .c and .h too. Signed-off-by: Florian Vaussard <[email protected]> Cc: Joe Perches <[email protected]> Acked-by: Rob Herring <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: fix spurious vendor compatible warningsFlorian Vaussard1-2/+2
With a compatible string like compatible = "foo"; checkpatch will currently try to find "foo" in vendor-prefixes.txt, which is wrong since the vendor prefix is empty in this specific case. Skip the vendor test if the compatible is not like compatible = "vendor,something"; Signed-off-by: Florian Vaussard <[email protected]> Cc: Joe Perches <[email protected]> Acked-by: Rob Herring <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: check vendor compatible with dashesFlorian Vaussard1-1/+1
The current vendor compatible check will not match vendors with dashes, like: compatible="asahi-kasei" Signed-off-by: Florian Vaussard <[email protected]> Reported-by: Joe Perches <[email protected]> Acked-by: Rob Herring <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: improve octal permissions test speedJoe Perches1-20/+31
The current octal permissions test is very slow. When patch ("checkpatch: add checks for constant non-octal permissions") was added, processing time approximately tripled. Regain almost all of the performance by not looping through all the possible functions unless the line contains one of the functions. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch.pl: modify warning message for printk usageYogesh Chaudhari1-1/+1
Modify warning message when printk is used in a patch. It mentions to use subsystem_dbg instead of netdev_dbg as the first preferred format of logging debug messages. Signed-off-by: Yogesh Chaudhari <[email protected]> Cc: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: make "return is not a function" test quieterJoe Perches1-7/+15
This test is a bit noisy and opinions seem to agree that it should not warn in a lot more situations. It seems people agree that: return (foo || bar); and return foo || bar; are both acceptable style and checkpatch should be silent about them. For now, it warns on parentheses around a simple constant or a single function or a ternary. return (foo); return (foo(bar)); return (foo ? bar : baz); The last ternary test may be quieted in the future. Modify the deparenthesize function to only strip balanced leading and trailing parentheses. Signed-off-by: Joe Perches <[email protected]> Cc: Julia Lawall <[email protected]> Reviewed-by: Josh Triplett <[email protected]> Cc: Monam Agarwal <[email protected]> Cc: Greg KH <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: ignore networking block comment style first lines in fileJoe Perches1-1/+2
It's very common to have normal block comments for the initial comments of a file description preface. So for files in drivers/net and net/ don't emit a warning when the first comment block in the file uses the normal block comment style and not the networking block comment style. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: use a more consistent function argument styleJoe Perches1-9/+20
Instead of array indexing $_, use temporary variables like all the other subroutines in the script use. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: add test for char * arrays that could be static constJoe Perches1-0/+7
static const char* arrays create smaller text as each function call does not have to populate the array. Emit a warning when char *arrays aren't static const and the array is not apparently global by being declared in the first column. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: fix jiffies comparison and othersJoe Perches1-1/+1
checkpatch could not distinguish between a variable in a struct named jiffies and the normal jiffies. foo->jiffies would emit a "Comparing jiffies" arning. Update the $Compare variable to do a negative look-behind for "-" when finding a ">" so that a pointer dereference like -> isn't a comparison. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: avoid sscanf test duplicated messagesJoe Perches1-1/+1
Change a test of $dstat to $line to avoid possibly emitting the sscanf warning multiple times. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: update octal permissions warningJoe Perches1-2/+4
When checking permissions, make sure 4 octal digits are used, but allow a single 0 too. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: warn on uses of __constant_<foo> functionsJoe Perches1-0/+13
Emit a warning when using any of these __constant_<foo> forms: __constant_cpu_to_be[x] __constant_cpu_to_le[x] __constant_be[x]_to_cpu __constant_le[x]_to_cpu __constant_htons __constant_ntohs Using any of these outside of include/uapi/ isn't preferred as using the function without __constant_ is identical when the argument is a constant. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: Simon Wunderlich <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: add checks for constant non-octal permissionsJoe Perches1-2/+34
umode_t permissions are sometimes mistakenly written with decimal constants. Verify that numeric permissions are using octal. Add a list of the most commonly used functions and macros that have umode_t permissions and the argument position. Add a $Octal type to $Constant. Allow $LvalOrFunc to be a pointer indirection too. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: don't warn on some function pointer return stylesJoe Perches1-10/+21
Checks for some function pointer return styles are too strict. Fix them. Multiple spaces after function pointer return types are allowed. int (*foo)(int bar) Spaces after function pointer returns of pointer types are not required. int *(*foo)(int bar) Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-04-03checkpatch: add test for long udelayJoe Perches1-3/+8
Holger reported: : The macro udelay cannot handle large values because of loss-of-precision. : : IMHO udelay on ARM is broken, because it also cannot work with fast : ARM processors (where bogomips >= 3355, which is in sight now). It's : just not broken enough that someone did something against it ... so : the current kludge is good enough. Until then, warn on long udelay uses. Also fix uses of $line that should have been $herecurr. Signed-off-by: Joe Perches <[email protected]> Reported-by: Holger Schurig <[email protected]> Cc: Sujith Manoharan <[email protected]> Cc: John Linville <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-02-10checkpatch: fix detection of git repositoryRichard Genoud1-2/+2
Since git v1.7.7, the .git directory can be a file when, for example, the kernel is a submodule of another git super project. So, the check "-d .git" is not working anymore in this case. Using a more generic check like "-e .git" corrects this behaviour. Signed-off-by: Richard Genoud <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-27checkpatch.pl: check for function declarations without argumentsJoe Perches1-0/+9
Functions like this one are evil: void foo() { ... } Because these functions allow variadic arguments without checking the arguments at all. Original patch by Richard Weinberger. Signed-off-by: Joe Perches <[email protected]> Cc: Richard Weinberger <[email protected]> Cc: Borislav Petkov <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: prefer ether_addr_copy to memcpy(foo, bar, ETH_ALEN)Joe Perches1-0/+10
ether_addr_copy was added for kernel version 3.14. It's slightly smaller/faster for some arches. Encourage its use. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: David Miller <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: add DT compatible string documentation checksRob Herring1-0/+27
This adds a simple check that any compatible strings in DeviceTree dts files are present in Documentation/devicetree/bindings. Vendor prefixes are also checked for existing in vendor-prefixes.txt These should be temporary checks until we have more sophisticated binding schema checking. Signed-off-by: Rob Herring <[email protected]> Signed-off-by: Joe Perches <[email protected]> Cc: Grant Likely <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: only flag FSF address, not gnu.org URLAlexander Duyck1-3/+1
This change restricts the check for the for the FSF address in the GPL copyright statement so that it only flags the address, not the references to the gnu.org/licenses URL which appears to be used in numerous drivers. The idea is to still allow some reference to an external copy of the GPL in the event that files are copied out of the kernel tree without the COPYING file. So for example this statement will still return an error: You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. However, this statement will not return an error after this patch: You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>. Signed-off-by: Alexander Duyck <[email protected]> Cc: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: add tests for function pointer style misusesJoe Perches1-0/+59
Kernel style uses function pointers in this form: "type (*funcptr)(args...)" Emit warnings when this function pointer form isn't used. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: Derek Perrin <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: update the FSF/GPL address checkJoe Perches1-5/+6
The FSF address check is a bit too verbose looking for the GPL text. Quiet it a bit by requiring --strict for the GPL bit. Also make the address tests match a few uses of abbreviations for street names and make it case insensitive. Signed-off-by: Joe Perches <[email protected]> Reviewed-by: Josh Triplett <[email protected]> Cc: Manfred Spraul <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: check for if's with unnecessary parenthesesJoe Perches1-0/+14
If statements don't need multiple parentheses around tested comparisons like "if ((foo == bar))". An == comparison maybe a sign of an intended assignment, so emit a slightly different message if so. Signed-off-by: Joe Perches <[email protected]> Reviewed-by: Josh Triplett <[email protected]> Cc: Manfred Spraul <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: improve space before tab --fix optionJoe Perches1-2/+4
This test should remove all the spaces before a tab not just one space. Substitute a tab for each 8 space block before a tab and remove less than 8 spaces before a tab. This SPACE_BEFORE_TAB test is done after CODE_INDENT. If there are spaces used at the beginning of a line that should be converted to tabs, please make sure that the CODE_INDENT test and conversion is done before this SPACE_BEFORE_TAB test and conversion. Reported-by: Manfred Spraul <[email protected]> Signed-off-by: Joe Perches <[email protected]> Cc: Josh Triplett <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: add a --fix-inplace optionJoe Perches1-1/+9
Add the ability to fix and overwrite existing files/patches instead of creating a new file "<filename>.EXPERIMENTAL-checkpatch-fixes". Suggested-by: Manfred Spraul <[email protected]> Signed-off-by: Joe Perches <[email protected]> Reviewed-by: Josh Triplett <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: attempt to find missing switch/case break;Joe Perches1-0/+25
switch case statements missing a break statement are an unfortunately common error. e.g.: commit 4a2c94c9b6c0 ("HID: kye: Add report fixup for Genius Manticore Keyboard") case blocks should end in a break/return/goto/continue. If a fall-through is used, it should have a comment showing that it is intentional. Ideally that comment should be something like: "/* fall-through */" Add a test to look for missing break statements. This looks only at the context lines before an inserted case so it's possible to have false positives when the context contains a close brace and the break is before the brace and not part of the patch context. Looking at recent patches, this is a pretty rare occurrence. The normal kernel style uses a break as the last line of the previous block. Signed-off-by: Joe Perches <[email protected]> Cc: Andy Whitcroft <[email protected]> Cc: Jiri Kosina <[email protected]> Cc: Benjamin Tissoires <[email protected]> Cc: Dave Jones <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: add warning of future __GFP_NOFAIL useDavid Rientjes1-0/+6
gfp.h and page_alloc.c already specify that __GFP_NOFAIL is deprecated and no new users should be added. Add a warning to checkpatch to catch this. Signed-off-by: David Rientjes <[email protected]> Cc: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: warn only on "space before semicolon" at end of lineJoe Perches1-1/+1
The "space before a non-naked semicolon" test has unwanted output when used in "for ( ;; )" loops. Make the test work only on end-of-line statement termination semicolons. Signed-off-by: Joe Perches <[email protected]> Cc: Dan Carpenter <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2014-01-23checkpatch: more comprehensive split strings warningJoe Perches1-8/+4
The current checkpatch test for split strings does not find several cases that should be found. For instance: /* Else poor success; go back to mode in "active" table */ } else { IWL_DEBUG_RATE(mvm, - "LQ: GOING BACK TO THE OLD TABLE suc=%d cur-tpt=%d old-tpt=%d\n", + "GOING BACK TO THE OLD TABLE: SR %d " + "cur-tpt %d old-tpt %d\n", window->success_ratio, window->average_tpt, lq_sta->last_tpt); does not currently emit a warning. Improve the test to find these cases. Add more exceptions to reduce false positives for assembly and octal/hex string constants. Signed-off-by: Joe Perches <[email protected]> Reviewed-by: Josh Triplett <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2013-12-13PCI/checkpatch: Deprecate DEFINE_PCI_DEVICE_TABLEJoe Perches1-4/+7
Prefer use of the direct definition of struct pci_device_id instead of indirection via macro DEFINE_PCI_DEVICE_TABLE. Update the PCI documentation to deprecate DEFINE_PCI_DEVICE_TABLE. Update checkpatch adding --fix option. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Bjorn Helgaas <[email protected]> Reviewed-by: Jingoo Han <[email protected]>
2013-11-21checkpatch: fix "Use of uninitialized value" warningsJoe Perches1-0/+1
checkpatch is currently confused about some complex macros and references undefined variables $stat and $cond. Make sure these are defined before using them. Signed-off-by: Joe Perches <[email protected]> Reported-by: Gerhard Sittig <[email protected]> Acked-by: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2013-11-13checkpatch: add check for sscanf without return useJoe Perches1-0/+17
Naked use sscanf can be troublesome because the pointed to variables may not have been set. Add a warning when the sscanf return value is not used. For now, do not add __must_check to the sscanf prototype because that will cause a couple of hundred new warnings when compiling a kernel. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2013-11-13checkpatch: don't require kernel style __attribute__ shortcuts in uapi pathsJoe Perches1-7/+12
Avoid prescribing kernel styled shortcuts for gcc extensions of __attribute__((foo)) in the uapi include paths. Fix $realfile filename when using -f/--file to not remove first level directory as if the filename was used in a -P1 patch. Only strip the first level directory (typically a or b) for P1 patches. Signed-off-by: Joe Perches <[email protected]> Cc: "Dixit, Ashutosh" <[email protected]> Cc: Andy Whitcroft <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2013-11-13checkpatch: improve "return is not a function" testJoe Perches1-14/+4
Find a few more cases where parentheses are used around the value of a return statement. This now uses the "$balanced_parens" test and also makes the test depend on perl v5.10 and higher. This now finds return with parenthesis uses the old code did not find like: ERROR: return is not a function, parentheses are not required #211: FILE: arch/m68k/include/asm/sun3xflop.h:211: + return ((error == 0) ? 0 : -1); Signed-off-by: Joe Perches <[email protected]> Tested-by: David Cohen <[email protected]> Acked-by: David Cohen <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.