- 29 Mar, 2017 40 commits
-
-
David Kershner authored
The only one using unisys_vmcall was visorchipset.c, it can be moved into the visorchipset file directly instead of being in a header file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The comment at the beginning of the vmcallinterface.h file references IO Virtualization. It should be more specific that it is referencing s-Par Virtualization. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The macro ISSUE_IO_VMCALL was a wrapper around the function unisys_vmcall. It doesn't need to exist and was just being noisy, so get rid of it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
We shouldn't be checking if VMCALL_CONTROLVM_ADDR is defined, remove it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
We had several if statements inside of if statements that should be consolidated into an if statement with an && to clean up the code some more. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Remove braces around single line if statements, they are not needed. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
If there is an error in chipset_device_pause/resume don't try to respond, error out and let the calling functions respond to this error just like any other error they encounter. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Clean up the function to remove some if statments that should be done in the calling function. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
If we fail to register the visordriver, don't call unregister, just return with the error. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't override the visorchannel_read_error, just return the error that it generates. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function dev_set_name can return an error, don't just ignore it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The add_uevent_var returns an error. Don't overwrite the error with -ENOMEM, just pass the error code back up. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The forcenomatch parameter was not being used, so get rid of it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The forcematch parameter was not being used, so get rid of it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Create bus_type was just calling register_bustype. Since we control how many times we call create_bus_type it was extraneous keeping a counter if the bus was already registered, so the functions were no longer needed. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function get_vbus_header_info returns errors that were being handled incorrectly. Fix the error handling in create_bus_instance. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Add error handling to visorbus_enable_channel_interrupt. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Report errors if we have a problem in dev_start_periodic_work. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
We only have one version of spar_check_channel, get rid of the client at the end and help save space. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The macro SPAR_VBUS_CHANNEL_CLIENT_OK was noisy and only used in one place so get rid of it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Move vbuschannel_print_devinfo from vbuchannel.h to visorbus_main.c since it is only referenced in visorbus_main.c. Since it is in a .c file, get rid of the inline for it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
With all the code churn, bus_responder and device_responder have become identical, consolidate them to the same function. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Adds error handling to the chipset_device_create message. If it returns a failure, it is assumed it has not signaled the s-Par firmware of the failure and the caller must do that. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Since signalinsert/remove now return valid error codes, we need to check them when we call them in visornic. The error codes need to propagate out to the calling functions. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Sameer Wadgaonkar authored
Added the __packed keyword to all structures in controlvmchannel.h to ensure alignment and size in memory. Signed-off-by: Sameer Wadgaonkar <sameer.wadgaonkar@unisys.com> Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function chipset_bus_create should return an error. If an error is returned, it is assumed the response has not been sent by bus_create. Correctly handle when the error has been returned in visorchipset. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The ret variable was only returning an error, so changing it to err to be more consistent across the file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The ret variable was only returning an error, so changing it to err to be more consistent across the file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The ret variable was only returning an error, so changing it to err to be more consistent across the file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The ret variable was only returning an error, so changing it to err to be more consistent across the file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The ret variable was only returning an error, so changing it to err to be more consistent across the file. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Removed a comment that was describing the relationship between the visorchipset driver and the visorbus driver. Since they are now one driver the comment no longer makes sense. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function bus_device_info_init was defined in visorbus_private.h. The only file that uses it is visorbus_main.c so move the function into the .c file instead. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't just drop the error from visorchannel_read on the floor, report it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't just drop the error from visorchannel_read on the floor, report it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't just drop the error from visorchannel_read on the floor, report it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't just drop the error from visorchannel_read on the floor, report it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
Don't just drop the error from visorchannel_read on the floor, report it. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function kobject_uevent_env can return an error, pass that error to the caller so it can be handled correctly. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
David Kershner authored
The function kobject_uevent_env returns an error we shouldn't just drop it on the floor but we should report it back to the caller. Since it now returns an error, need to add proper error handling to parahotplug_process_message. Signed-off-by: David Kershner <david.kershner@unisys.com> Reviewed-by: Reviewed-by: Tim Sell <timothy.sell@unisys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-