- Sep 04, 2012
-
-
Federico Vaga authored
Do not use device->id attribute but define our zio_device->dev_id attribute. This allow ZIO to run on kernel older than 3.2 Signed-off-by:
Federico Vaga <federico.vaga@gmail.com>
-
- Aug 29, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com>
-
- Aug 21, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com>
-
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
The core now use "hw" instead of "hwdev" as prefix for the fake zio_device, and no-prefix for the real zio_device. Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
The biggest change (as for number of lines) is renaming he ztt_instance structure pointers to ztt. Also, remove a trailing semicolon in a #define, fix a kfree call and remove an old comment that doesn't apply any more. Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
This adds 4 bytes to the control (there are atill 16 free) to fit a whole struct addr_zio (same as sockaddr_zio, to be defined when we merge PF_ZIO). The structure was almost ready for that, but we missed the sa_family 16-bit field. This also changes the minor number for the ZIO control structure, and all the few users of the address-related fields. Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
- Aug 10, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
A new zio_attribute flag is declared: ZATTR_CONTROL. If this flag is setted the zio_attribute appear in the control, otherwise the attribute is visibile in the sysfs but not in the control Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
The ZIO core use the dev_id parameter but this patch is compatible only with kernel 3.2 or greater. Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
- May 03, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
- May 02, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
- Apr 17, 2012
-
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Federico Vaga authored
The helper zio_trigger_abort() is introduced; it can be used to abort a data transfer. The bit flag ZTI_COMPLETING is introduced: if high ZIO is completing the transfert operation by invokng the data_done() function; if low the transfer is running or the trigger is waiting for event. If abort() is called when this glag is high, the functions does nothing. The helper zio_trigger_data_done() is introduced; insted of call the data_done() function with cset->ti->t_op->data_done() and manually set the ZTI_COMPLETING flag, this hellper must be used. If the trigger operation data_done is not set, then use the __zio_internal_data_done Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
If default_(zbuf|trig) is not empty then ZIO use this as first choice for the buffer|trigger; otherwise it tries to use the preferred_(buffer|trigger) defined at zio_device and valid for any cset. The last choice is the ZIO default. Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
- Apr 16, 2012
-
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
- Apr 09, 2012
-
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Alessandro Rubini authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Alessandro Rubini authored
The core should ensure its users that create calls are not reentrant. We use the locks that already exist in the buffer type and the trigger type. Then, we init both ti->lock and bi->lock, so our users can rely on them (see next patch). Related to locks: fix a comment in zio.h Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
- Apr 06, 2012
-
-
Alessandro Rubini authored
Signed-off-by:
Alessandro Rubini <rubini@gnudd.com> Acked-by:
Federico Vaga <federico.vaga@gmail.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
This patch introduces zio_driver and zio_device_id structure, which are used to define a zio's driver. ZIO uses these structures to handle the match between devices and drivers connected on the zio bus. To make ZIO work on the Linux bus mechanism, ZIO creates two zio_device structures. The first zio_device is allocated and it is registered by the module which wants to declare the existence of a new ZIO device. This first zio_device is only used by ZIO to match the device with a driver. When the first zio device matchs with a driver, ZIO creates a new zio device child of the first one. The new device is filled with all the device information and sysfs attributes. This second zio device represents the real device to use; ZIO passes this zio device as argument to the driver probe function, so the driver can use the real zio device for its operations. From the user space point of view, both ZIO devices are available at /sys/bus/zio/devices/; the first zio device has the prefix "hwdev-", the second zio device (the real one) has the prefix "zio-" Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
cset and bi use a different set of sysfs default attributes, so now they use their own device_type with their default attributes Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com> Acked-by:
Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by:
Federico Vaga <federico.vaga@gmail.com>
-