- 09 Jul, 2014 5 commits
-
-
Sebastian Ott authored
GenWQE used to call pci_enable_msi_block to allocate a desired number of MSI's. If that was not possible pci_enable_msi_block returned with a smaller number which might be possible to allocate. GenWQE then called pci_enable_msi_block with that number. Since commit a30d0108 "GenWQE: Use pci_enable_msi_exact() instead of pci_enable_msi_block()" pci_enable_msi_exact is used which fails if the desired number of MSI's was not possible to allocate. Change GenWQE to use pci_enable_msi_range to restore the old behavior. Signed-off-by:
Sebastian Ott <sebott@linux.vnet.ibm.com> Reviewed-by:
Alexander Gordeev <agordeev@redhat.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Kleber Sacilotto de Souza authored
Increase genwqe driver version number. Signed-off-by:
Kleber Sacilotto de Souza <klebers@linux.vnet.ibm.com> Acked-by:
Frank Haverkamp <haver@linux.vnet.ibm.com> Signed-off-by:
-
Kleber Sacilotto de Souza authored
Currently, in the event of a fatal hardware error, the driver tries a recovery procedure that calls pci_reset_function() to reset the card. This is not sufficient in some cases, needing a fundamental reset to bring the card back. This patch implements a call to the platform fundamental reset procedure on the error recovery path if GENWQE_PLATFORM_ERROR_RECOVERY is enabled. This is implemented by default only on PPC64, since this can cause problems on other archs, e.g. zSeries, where the platform has its own recovery procedures, leading to a potencial race conditition. For these cases, the recovery is kept as it was before. Signed-off-by:
-
Kleber Sacilotto de Souza authored
This patch implements the callbacks and functions necessary to have EEH recovery support. It adds a config option to enable or disable explicit calls to trigger platform specific mechanisms on error recovery paths. This option is enabled by default only on PPC64 systems and can be overritten via debugfs. If this option is enabled, on the error recovery path the driver will call pci_channel_offline() to check for error condition and issue non-raw MMIO reads to trigger early EEH detection in case of hardware failures. This is necessary since the driver MMIO helper funtions use raw accessors. Signed-off-by:
-
Kleber Sacilotto de Souza authored
This patch adds an interface on sysfs for userspace to request a card bitstream reload. It sets the appropriate register and try to perform a fundamental reset on the PCIe slot for the card to reload the bitstream from the chosen partition. Signed-off-by:
-
- 08 Jul, 2014 1 commit
-
-
Himangi Saraogi authored
nhi->rx_rings does not have type as struct tb_ring *, as it is a double pointer so the elements of the array should have pointer type, not structure type. The Coccinelle semantic patch that makes this change is as follows: // <smpl> @disable sizeof_type_expr@ type T; T **x; @@ x = <+...sizeof( - T + *x )...+> // </smpl> Signed-off-by:
Himangi Saraogi <himangi774@gmail.com> Acked-by:
Julia Lawall <julia.lawall@lip6.fr> Cc: Andreas Noever <andreas.noever@gmail.com> Signed-off-by:
-
- 23 Jun, 2014 1 commit
-
-
Arnd Bergmann authored
commit 94859308 "w1: new w1_ds2406 driver" added a new driver that uses the crc16 library, but didn't ensure that the core is there. This adds the necessary Kconfig statements, just like we have it for other w1 drivers. Reported-by:
kbuild test robot <fengguang.wu@intel.com> Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Acked-by:
Scott Alfter <scott@alfter.us> Cc: Evgeniy Polyakov <zbr@ioremap.net> Signed-off-by:
-
- 21 Jun, 2014 1 commit
-
-
Andreas Noever authored
We use __crc32c_le in ctl.c. So make sure that the dependency is there. Reported-by:
Andreas Noever <andreas.noever@gmail.com> Signed-off-by:
-
- 20 Jun, 2014 21 commits
-
-
Andreas Noever authored
tb_eeprom_get_drom_offset is local to this file. Signed-off-by:
-
Andreas Noever authored
Force enum tb_drom_entry_type to unsigned to fix the following error: drivers/thunderbolt/eeprom.c:202:39: error: dubious one-bit signed bitfield Signed-off-by:
-
Andreas Noever authored
tb_find_cap in cap.c takes an enum tb_cap and not an u32. Fix the declaration in tb.h. Signed-off-by:
-
Andreas Noever authored
Thunderbolt packets are big endian. Cast pkg->buffer to __be32* when accessing the checksum. Reported-by:
-
Arnd Bergmann authored
The result of "sizeof(struct tb_drom_entry_port)" is a size_t, which is not necessarily the same as 'long', so we should use the appropriate %z format string instead of %l. Signed-off-by:
-
Arnd Bergmann authored
The thunderbolt drivers cannot be built if CONFIG_PCI is disabled, better add an explicit Kconfig dependency. The "default no" line is redundant and can be removed at the same time. Signed-off-by:
-
Sachin Kamat authored
'nhi_ids' is local to this file. Signed-off-by:
Sachin Kamat <sachin.kamat@samsung.com> Acked-by:
-
Sachin Kamat authored
'descriptors' is a pointer. Use NULL isntead of 0. Signed-off-by:
-
Sachin Kamat authored
The function returns a pointer. Hence return NULL instead of 0. Signed-off-by:
-
Sachin Kamat authored
The function returns a pointer. Hence return NULL instead of 0. Signed-off-by:
-
Sachin Kamat authored
Fixes the below error: drivers/thunderbolt/switch.c:347:2: error: implicit declaration of function ‘kzalloc’ [-Werror=implicit-function-declaration] drivers/thunderbolt/switch.c:381:2: error: implicit declaration of function ‘kcalloc’ [-Werror=implicit-function-declaration] Signed-off-by:
-
Sachin Kamat authored
Fixes the below error: drivers/thunderbolt/eeprom.c:407:2: error: implicit declaration of function ‘kzalloc’ [-Werror=implicit-function-declaration] drivers/thunderbolt/eeprom.c:444:2: error: implicit declaration of function ‘kfree’ [-Werror=implicit-function-declaration] Signed-off-by:
-
Fjodor Schelichow authored
This patch replaces all calls to the "printk" function within the "masters" subdirectory by calls to the appropriate "pr_*" function thus addressing the following warning generated by the checkpatch script: WARNING: Prefer [subsystem eg: netdev]_err([subsystem]dev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ... Signed-off-by:Fjodor Schelichow <fjodor.schelichow@hotmail.com> Signed-off-by:
Roman Sommer <romsom2@yahoo.de> Acked-by:
Evgeniy Polyakov <zbr@ioremap.net> Signed-off-by:
-
Fjodor Schelichow authored
This patch replaces all calls to the "printk" function within the main "w1" directory by calls to the appropriate "pr_*" function thus addressing the following warning generated by the checkpatch script: WARNING: Prefer [subsystem eg: netdev]_err([subsystem]dev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ... Signed-off-by: -
Fjodor Schelichow authored
This patch replaces all calls to the "printk" function within the "slaves" subdirectory by calls to the appropriate "pr_*" function thus addressing the following warning generated by the checkpatch script: WARNING: Prefer [subsystem eg: netdev]_err([subsystem]dev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ... Signed-off-by: -
Scott Alfter authored
Some preliminary work at making use of this driver led me to implement CRC-16 checks on read and write to deal with the occasional glitchiness of the 1-Wire bus. The revised driver (attached) returns an I/O error if the CRC check fails. When reading the chip's state, either you get a valid indication or you get an I/O error. When changing its state, either the change is successful or an I/O error is returned. Signed-off-by:
-
Thomas Wood authored
Replace spaces at beginning of the string with tabs, and replace foo * bar with foo *bar in a pointer declaration. Signed-off-by:
Thomas Wood <tommyandrena@gmail.com> Acked-by:
-
Alexander Shiyan authored
According to the i.MX reference manual, the read/write bit operations takes from 60 us to 120 us. This patch optimizes mxc_w1_ds2_touch_bit() function to use proper value for such delay. Nevertheless, a small margin for the timeout has been added for the case if clock frequency is inaccurate. Signed-off-by:
Alexander Shiyan <shc_work@mail.ru> Signed-off-by:
-
Alexander Shiyan authored
This patch adds a software reset for 1-Wire module at driver startup. Signed-off-by:
-
Alexander Shiyan authored
According to the i.MX reference manual, the reset procedure and "presence" pulse takes 511 and 512 us, respectively. Measurement for i.MX27 is about 1100 us. There is no need to wait Reset+Presence more than this time. This patch optimizes mxc_w1_ds2_reset_bus() function to use proper value for delay after w1 bus reset. Nevertheless, a small margin for the timeout has been added for the case if clock frequency is inaccurate. Signed-off-by:
-
Alexander Shiyan authored
W1 reset_bus() should return zero if slave device is present. This patch fix this issue. Signed-off-by:
-
- 19 Jun, 2014 11 commits
-
-
Andreas Noever authored
Fix issues observed with the Startech docking station: Fix the type of the route parameter in tb_ctl_rx. It should be u64 and not u8 (which only worked for short routes). A thunderbolt cable contains two lanes. If both endpoints support it a connection will be established on both lanes. Previously we tried to scan below both "dual link ports". Use the information extracted from the drom to only scan behind ports with lane_nr == 0. Endpoints with more complex thunderbolt controllers have some of their ports disabled (for example the NHI port or one of the HDMI/DP ports). Accessing them results in an error so we now ignore ports which are marked as disabled in the drom. Signed-off-by:
-
Andreas Noever authored
All Thunderbolt switches (except the root switch) contain a drom which contains information about the device. Right now we only read the UID. Add code to read and parse this drom. For now we are only interested in which ports are disabled and which ports are "dual link ports" (a physical thunderbolt port/socket contains two such ports). Signed-off-by:
-
Andreas Noever authored
We use _noirq since we have to restore the pci tunnels before the pci core wakes the tunneled devices. Signed-off-by:
-
Andreas Noever authored
Add eeprom access code and read the uid during switch initialization. The UID will be used to check device identity after suspend. Signed-off-by:
-
Andreas Noever authored
Add two quirks to support thunderbolt suspend/resume on Apple systems. We need to perform two different actions during suspend and resume: The whole controller has to be powered down before suspend. If this is not done then the native host interface device will be gone after resume if a thunderbolt device was plugged in before suspending. The controller represents itself as multiple PCI devices/bridges. To power it down we hook into the upstream bridge of the controller and call the magic ACPI methods. Power will be restored automatically during resume (by the firmware presumably). During resume we have to wait for the native host interface to reestablish all pci tunnels. Since there is no parent-child relationship between the NHI and the bridges we have to explicitly wait for them using device_pm_wait_for_dev. We do this in the resume_noirq phase of the downstream bridges of the controller (which lead into the thunderbolt tunnels). Signed-off-by:
-
Andreas Noever authored
Add pci_fixup_suspend_late as a new pci_fixup_pass. The pass is called from suspend_noirq and poweroff_noirq. Using the same pass for suspend and hibernate is consistent with resume_early which is called by resume_noirq and restore_noirq. The new quirk pass is required for Thunderbolt support on Apple hardware. Signed-off-by:
Bjorn Helgaas <bhelgaas@google.com> Signed-off-by:
-
Andreas Noever authored
A pci downstream and pci upstream port can be connected through a tunnel. To establish the tunnel we have to setup two unidirectional paths between the two ports. Right now we only support paths with two hops (i.e. no chaining) and at most one pci device per thunderbolt device. Signed-off-by:
-
Andreas Noever authored
A thunderbolt path is a unidirectional channel between two thunderbolt ports. Two such paths are needed to establish a pci tunnel. This patch introduces struct tb_path as well as a set of tb_path_* methods which are used to activate & deactivate paths. Signed-off-by:
-
Andreas Noever authored
We receive a plug event callback whenever a thunderbolt device is added or removed. This patch fills in the tb_handle_hotplug method and starts reacting to these events by adding/removing switches from the hierarchy. Signed-off-by:
-
Andreas Noever authored
Add utility methods tb_port_state and tb_wait_for_port. Add tb_scan_switch which recursively checks for downstream switches. Signed-off-by:
-
Andreas Noever authored
Thunderbolt switches have a plug events capability. This patch adds the tb_plug_events_active method and uses it to activate plug events during switch allocation. Signed-off-by:
-
