Provide common routine for the transition of operational state for a leaf
device during a root device transition.
Signed-off-by: Patrick Mullaney <[email protected]>
Acked-by: Arnd Bergmann <[email protected]>
Acked-by: Patrick McHardy <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
- Replace magic values with constants
- Simplify length calculation and fix a bug
Based on valuable feedback from Ben Hutchings <[email protected]>
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
when disabling interrupts, driver was writing with IO, this is no
necessary because on ixgbe parts the hardware can "oneshot"
disable and clear the interrupt. So on 82598/82599 use of EIAM
should avoid one posted write per interrupt when in MSI-X mode.
This should improve performance and seems to in my limited
testing, reduce CPU utilization VERY slightly.
Signed-off-by: Jesse Brandeburg <[email protected]>
Acked-by: Peter P Waskiewicz Jr <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
drop variables that had cache lines modified in simultaneous hot paths.
keep some variables modified on hot paths but make their storage per queue.
cache align DMA data buffer start addresses.
cache align (padding) some structures that end within a cacheline.
Signed-off-by: Jesse Brandeburg <[email protected]>
Acked-by: Peter P Waskiewicz Jr <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
decrease the memory utilization of the tx / rx queue allocation
by changing the default ring size to 512 (from 1024). At
1024 rx entries of 2KB each (from 4kB slab) with 16 queues
ixgbe was using 64 MB of memory per port, which is not
necessary.
Users can still change queue lengths with ethtool -k.
Signed-off-by: Jesse Brandeburg <[email protected]>
Acked-by: Peter P Waskiewicz Jr <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This removes the Tx queue selection for FCoE traffic from ixgbe_xmit_frame()
and does it in the ndo_select_queue() call, moving all Tx queue selection
into a single routine.
Signed-off-by: Yi Zou <[email protected]>
Acked-by: Peter P Waskiewicz Jr <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Store the user priority for FCoE and use it directly for outgoing
FCoE traffic when DCB is enabled.
Signed-off-by: Yi Zou <[email protected]>
Acked-by: Peter P Waskiewicz Jr <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Using remote wakeup and delayed transmission to allow
online device to go into usb autosuspend.
Minimal alternate support for devices that don't support
remote wakeup.
Signed-off-by: Oliver Neukum <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Normal resume can do double duty as reset_resume() for this driver
as it keeps no state in the device
Signed-off-by: Oliver Neukum <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
ata_set_lba_range_entries used the variable max for two different things
which was confusing. Make the function take a buffer size in bytes as
argument and return the used buffer size upon completion.
Signed-off-by: Martin K. Petersen <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Our current TRIM payload is a single sector that can accommodate 64 *
65535 blocks being unmapped. Report this value in the Block Limits
Maximum Unmap LBA count field.
If a storage device supports TRIM and the DRAT and RZAT bits are set,
report TPRZ=1 in Read Capacity(16).
Signed-off-by: Martin K. Petersen <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Remove superfluous timings table entry while at it.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Acked-by: Sergei Shtylyov <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
MWDMA0 timings cannot be met with the PIIX based controller
programming interface.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Acked-by: Sergei Shtylyov <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
The naming of the defines suggests that there are three IISv4 ports
with one data line each when in fact there is a single IISv4 port
with three data lines.
Signed-off-by: Mark Brown <[email protected]>
Signed-off-by: Ben Dooks <[email protected]>
Tg3's firmware is no longer loaded from tg3_init_one(). Remove the
error path that releases the firmware back to the system.
Signed-off-by: Matt Carlson <[email protected]>
Reviewed-by: Michael Chan <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
If the driver needs to use PCI configuration space to read the VPD data,
the accesses could collide with the pci_read_vpd() function. This patch
converts the driver to use pci_read_vpd() to avoid any potential
collisions.
Signed-off-by: Matt Carlson <[email protected]>
Reviewed-by: Michael Chan <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This patch cleans up the VPD code by creating preprocessor definitions
and using them in the place of hardcoded constants.
Signed-off-by: Matt Carlson <[email protected]>
Reviewed-by: Michael Chan <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The 57765 asic rev has MSI-X capability, but does not support TSS. This
patch changes the tx paths so that TSS is explicitly mentioned, rather
than implied through the ENABLE_MSIX flag.
Signed-off-by: Matt Carlson <[email protected]>
Signed-off-by: Michael Chan <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Only files where David Miller is the primary git-signer.
wireless, wimax, ixgbe, etc are not modified.
Compile tested x86 allyesconfig only
Not all files compiled (not x86 compatible)
Added a few > 80 column lines, which I ignored.
Existing checkpatch complaints ignored.
Signed-off-by: Joe Perches <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Set correct bits to switch between UDMA2 and UDMA4.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Fix ->resume method to do chipset specific setup.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
We need to mask out the port offset from the port number
cached in ld_qdi->timing.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Dual port QDI6580 has shared PIO timings for master/slave
devices so it needs to use custom ->qc_issue method.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Clear old MWDMA timings before programming new ones (IT8213
is a single port host so there is no need to check ap->port_no).
This change should be safe as this is how we have been doing
things in IDE it8213 host driver for years.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
[ port of Sergei's fixes for pata_efar from commit 5f33b3b ]
Fix the PIO mode 2 using mode 0 timings -- this driver should enable the
fast timing bank starting with PIO2, just like the PIIX/ICH drivers do.
Also, fix/rephrase some comments while at it.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
* do not clear PIO timings for master when programming slave
* program new PIO timings in the correct register nibble
Both changes should be safe as this is how we have been doing
things in IDE it8213 host driver for years.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Driver should program the cycle timing not the mode number
(doing the latter results in wrong timings being used).
There shouldn't be any problems with it as IDE it8213 host driver
has been doing it this way for years.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
It has been dead for the last three years (== since the initial driver
merge) and probability that it will ever get fixed is quite low.
Since there is no reason to keep this dead code around any longer just
remove it (it can still be retrieved from the git history if necessary).
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
Do not clear MWDMA timings for device on the other port when
programming slave device.
This change should be safe as this is how we have been doing
things in IDE slc90e66 host driver for years.
Signed-off-by: Bartlomiej Zolnierkiewicz <[email protected]>
Signed-off-by: Jeff Garzik <[email protected]>
