- 14 Apr, 2017 18 commits
-
-
Colin Ian King authored
char array cmd is being scanned in using a %4s scanf format specifier and so cmd must be an array of 5 chars. Increase size to 5 chars to ensure we don't have an overflow. Detected with static analysis by cppcheck: "(error) Width 4 given in format string (no. 1) is larger than destination buffer 'cmd[4]', use %3s to prevent overflowing it." Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Dan Carpenter authored
lustre_cfg_new() returns error pointers and never NULL. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Valerio Genovese authored
Removed extra space before comma in memset() as a part of checkpatch.pl fix-up. Signed-off-by: Valerio Genovese <valerio.click@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Arnd Bergmann authored
In combination with CONFIG_PROFILE_ANNOTATED_BRANCHES=y, the unlikely() inside of the WARN() macro becomes too complex for gcc to see that we don't use the output arguments of mt9m114_to_res() are used correctly: drivers/staging/media/atomisp/i2c/mt9m114.c: In function 'mt9m114_get_fmt': drivers/staging/media/atomisp/i2c/mt9m114.c:817:13: error: 'height' may be used uninitialized in this function [-Werror=maybe-uninitialized] int width, height; ^~~~~~ drivers/staging/media/atomisp/i2c/mt9m114.c: In function 'mt9m114_s_exposure_selection': drivers/staging/media/atomisp/i2c/mt9m114.c:1179:13: error: 'height' may be used uninitialized in this function [-Werror=maybe-uninitialized] Without WARN_ON(), there is no problem, so by simply replacing it with v4l2_err(), the warnings go away. The WARN() output is also not needed here, as we'd probably catch the problem before even getting here, and other checks for the same condition already use v4l2_err. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
UDS is another layer which actually boils down to some trivial assignments so remove it so inline the code. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
These are not used in the driver so can go away. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
This is a whole pile of code that wraps a single assignment. Remove it and put the assignment in the caller. Once we have the kernels sorted we should revisit these and remove all the pointless 1 item structs that go with it. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
This isn't used so it can go in the bitbucket. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
The base hmm MMU code doesn't support contiguous allocations (they BUG), so remove support from them from the higher levels of the heirarchy. We still need to unwind all these layers but it turns out that some of the init order stuff is rather sensitive and the simple cleanup breaks everything Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
Where we know the buffer size is reasonably constrained we can just use kmalloc, and where it will be large vmalloc. This still leaves a pile in the middle. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
We have one hard coded set of behaviour so unpick the indirection and function pointers. This isn't the whole story. A lot of the callers are known sizes and use cases so we can switch them directly to kmalloc later on. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
We are not using these for debugging or debug logging so remove the defines, trim and rename the functions. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Alan Cox authored
They can be replaced by kmalloc. There are a few that do need to pick kmalloc or vmalloc. Those we leave for the moment. Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Daeseok Youn authored
The line which is initializing mipi_info variable is too long to read. It would be placed in next line. Signed-off-by: Daeseok Youn <daeseok.youn@gmail.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Daeseok Youn authored
The address of isp->asd[i] is already assigned to local "asd" variable. "&isp->asd[i]" would be replaced with just "asd". Signed-off-by: Daeseok Youn <daeseok.youn@gmail.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Daeseok Youn authored
Enable/Disable ISP irq is switched with "enable" parameter of enable_isp_irq(). It would be better splited to two such as enable_isp_irq()/disable_isp_irq(). But the enable_isp_irq() is no use in atomisp_cmd.c file. So remove the enable_isp_irq() function and add disable_isp_irq function only. Signed-off-by: Daeseok Youn <daeseok.youn@gmail.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Colin Ian King authored
trivial fix to spelling mistake in dev_err error message Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Daeseok Youn authored
Define new local variable to reduce the number of reference. The new local variable is added to save the addess of dfs and used in atomisp_freq_scaling() function. Signed-off-by: Daeseok Youn <daeseok.youn@gmail.com> Signed-off-by: Alan Cox <alan@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
- 11 Apr, 2017 22 commits
-
-
Tobin C. Harding authored
Function uses an if statement within a for loop to guard a block of code. If 'if' statement conditional evaluates to false, loop breaks. The same logic can be expressed by inverting the conditional and breaking when new conditional evaluates to true. This allows the subsequent code to be indented one level less, aiding readability. Reduced indentation also allows for the code to be laid out more clearly and fixes two checkpatch warnings. Invert if statement conditional, break from for loop if new conditional evaluates to true. Reduce indentation in subsequent code, fix whitespace issues. Do not change program logic. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function uses multi-way decision for control flow. Final statement of function is spin_unlock(). Code can be simplified by adding a goto label labelling the call to spin_unlock() and jumping to label instead of using multi-way decision. This allows the code to be indented one level less which adds to the readability. Add goto label. Remove multi-way decision by jumping to label. Reduce indentation in subsequent code. Do not change program logic. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Macro CHECK_ALINE is defined and never used. Remove unused macro. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Checkpatch emits WARNING: Avoid multiple line dereference. Fix up layout of function call, move dereference to single line. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Checkpatch emits multiple CHECK: Logical continuations should be on the previous line. Move logical continuations to the end of the previous line. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Kernel networking code predominately uses the identifier 'skb' for a struct sk_buff pointer. Of 8088 instances of 'struct sk_buff *' within net/ 6670 are named 'skb'. Following the principle of least surprise, new networking code should use the identifier 'skb' for variables of type 'struct sk_buff *'. Rename identifier 'packet' to 'skb'. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Driver uses custom Michael MIC implementation. There is already an implementation within the kernel. There is at least one other driver already using the kernel implementation (drivers/net/wireless/intersil/orinoco). Add task to TODO file. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Identifier uses camel case, standard kernel style does not use camel case. Rename buffer 'RecvMIC' to 'recv_mic'. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Checkpatch emits WARNING: else is not generally useful after a break or return. Two warnings of this type are emitted for this code block, in both cases 'else' statements are unnecessary. Remove unnecessary 'else' statements, reduce indentation in subsequent code. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Checkpatch emits CHECK: Alignment should match open parenthesis. Align argument to open parenthesis. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function contains a local pointer variable defined to a memory location within a structure. This memory location is later used by dereferencing the struct instead of using the local pointer. The code is cleaner if all references of the same memory location use the local variable. Utilize existing local pointer variable instead of dereferencing struct. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function uses overly complex calls to memcpy(). Code may be simplified by the use of a local variable. Code sometimes uses explicit address of initial array element and sometimes does not. Uniformity aids readability. If array pointers are explicit it aids readability further. Simplify calls to memcpy(). Add local pointer variable, define it to the correct memory location. Use newly defined variable in calls to memcpy(). Be uniform in use of explicit address of first element of array (&foo[0]). Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function parameter is cast to a local pointer which is then dereferenced before it is checked to be non-NULL. Move pointer null check to be before the pointer is dereferenced. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Multi-way decision contains two anomalies. Firstly, a local variable is defined to be the inverse truth variable of a struct member. This local variable is used as the conditional to the multi-way decision. This is unnecessary, the same logic can be expressed using the struct member directly. Secondly, there are four branches in the multi-way decision, two of which can never be executed. This is dead code. Remove unnecessary local variable. Remove two branches of multi-way decision statement that can never be executed. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function contains compound statement delineated by lone braces. This statement represents a discreet set of functionality and thus can be factored out into a separate function. Using a separate function instead of a compound statement increases readability, reduces code indentation, reduces function length, and generally looks more tidy. Factor compound statement out to separate function. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function has duplicate code clean up sequences; identical function call followed by return. This would be better expressed with the use of a goto statement, as is typical in-tree. One call site places the clean up code within the 'else' branch of an multi-way decision. This can be more clearly expressed by inverting the initial decision conditional and jumping directly to the cleanup code. Subsequent code indentation can then be reduced, aiding readability. Fix function return code execution path. Move clean up code to end of function with a label. Replace duplicate clean up code within function with a jump to label. Invert conditional, jump to label if new conditional evaluates to true, reduce indentation in subsequent code. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function contains a list of four checks, for no apparent reason two of them are OR'ed together. Having two OR'ed together and the other two not implies some connection between the two that are combined. It is easier to read this code if the four unrelated checks are done as separate statements. Move dissimilar expressions out of logical continuation and into separate statement. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function has return type 'int'. Function has internal linkage. Function returns 0 on all execution paths. Function is called only once in the driver and the return value is not checked. Removal of this return value does not change the program logic. The 'int' return type is not adding any information thus it is better to remove it. Change return type of function with internal linkage from 'int' to 'void'. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
A number of functions have blocks of code guarded by an if statement. if (foo) { /* block of code */ } This can, on occasion, more succinctly be expressed as if (!foo) return /* block of code */ This change will mean a number of whitespace issues need to be addressed/fixed. The diff can be a little hard to read when there are multiple lines that are very similar (for example error return code). Splitting the indentation reduction and the whitespace fixes into two separate patches was not found to aid reading the diff. Therefor we fix the whitespace issues at the same time. We need to be very sure to not make any changes to the program logic, this is ensured by only doing what is stated - invert the conditional, fix whitespace. Invert if statement conditionals. Return if new conditional evaluates to true. Reduce level of indentation in subsequent code. Fix white space issues. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Function uses identifier 'retval' to hold the error return value. The rest of the driver uses 'ret' for this purpose. Being uniform in the choice of identifiers generally adds to the cleanliness of the code, also it is arguably easier to follow the code if one name is used for one task. Rename identifier 'retval' to 'ret'. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Driver uses identifier 'rc' to hold the value for error return code. The rest of the driver predominately uses 'ret' for this purpose. It is easier to follow the code if one name is used for one task. Rename identifier 'rc' to 'ret'. Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Tobin C. Harding authored
Length undergoes type conversion when passed (indirectly) as an argument for parameter of type 'unsigned int'. If length is negative this is a bug (the value after conversion is large). Declare 'length' to be an unsigned type instead of a signed type. Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tobin C. Harding <me@tobin.cc> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-