| Commit message (Collapse) | Author | Age |
|---|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* 'viafb-next' of git://github.com/schandinat/linux-2.6: (24 commits)
viafb: Automatic OLPC XO-1.5 configuration
viafb: remove unused CEA mode
viafb: try to map less memory in case of failure
viafb: use write combining for video ram
viafb: add X server compatibility mode
viafb: reduce OLPC refresh a bit
viafb: fix OLPC XO 1.5 device connection
viafb: fix OLPC DCON refresh rate
viafb: delete clock and PLL initialization
viafb: replace custom return values
viafb: some small cleanup for global variables
viafb: gather common good, old VGA initialization in one place
viafb: add engine clock support
viafb: add VIA slapping capability
viafb: split clock and PLL code to an extra file
viafb: add primary/secondary clock on/off switches
viafb: add clock source selection and PLL power management support
viafb: prepare for PLL separation
viafb: call viafb_get_clk_value only in viafb_set_vclock
viafb: remove unused max_hres/vres
...
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Currently, a long set of viafb options are needed to get the XO-1.5
laptop to output video (there is only 1 configuration that works, that
can't really be autodetected).
This patch automatically detects and configures viafb for the XO-1.5
laptop, meaning all that is required for working display is that
viafb is loaded.
Signed-off-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |
| |
| |
| |
| |
| | |
This trivial patch removes unused mode tables.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The current code might result in trying to remap 512MB video ram on
a 32 bit system which is quite likely to fail.
This patch tries to map less of it down to 8MB as this should still
be enough to get a reasonably well working framebuffer. This should
make viafb work for many people without requiring them to manually
allocate more space.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |
| |
| |
| |
| |
| | |
This can give a speed up of factor 6-9, which is quite notable.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |\
| | |
| | |
| | |
| | |
| | |
| | | |
Conflicts:
drivers/video/via/viamode.c
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch adds a config option to be compatible with X servers like
OpenChrome. This is required as for example the X server does not
handle things like disabled IGAs/PLLs resulting in a potential
freeze on X startup. With this option disabled we can provide some
nice features like power management and not reinitializing the
hardware on every mode switch (taking long time, causing flickering).
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
We do this also in the real program code so there is no reason to
do it here too (and here it's hardly readable).
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch adds support for enabling and configuring the engine on
VIAs IGPs. This is the main clock used for everything but pixel
output.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch introduces dummy functions to execute when we don't know
what we should do (due to missing documentation). They do nothing
but print a nice message in the log explaining the situation.
To trigger this message initial power management support is
activated which might save a bit energy by disabling PLL and clock
if no device is configured to use them.
Note: The message is only shown for the oldest IGPs CLE266 and K400
as for the other platforms there are reasonable assumptions
how it does (hopefully) work.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch is a huge move operation with some rename and introduces
an interface to still use the functions. This should be a step in
the right direction to reuse the code whenever possible but cleanly
separate code that differs on different platform and keeping the
complexity as low as possible.
pll_config was renamed to via_pll_config to keep the naming scheme.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch adds functions to enable/disable the display clocks.
It also fixes a tiny bug that slipped in with a previous commit but
could not yet have caused any problems.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch adds some support for clock source selection as well as
PLL power management. The code is unused at the moment but was
successfully tested as far as possible.
The implementation is according to the documentation for VX700,
VX800, VX855, VX900. Probably the source selection works like this
starting with K800 and the power managemennt at least since VX700.
(guessed based on the initialization in viamode.c)
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | | |
This patch splits some functionality to extra functions.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
As no caller is interested in the result call viafb_get_clk_value
directly from viafb_set_vclock to encapsulate the hardware dependend
stuff there.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch allows calculating the pll multiplier within limits based
on the previous table. All available information supports that it
should be possible/sane to choose the multiplier free within some
ranges.
Storing the multiplier ranges instead of lots of pll configurations
reduces the memory needed and may as well improve the performance.
It is also expected to provide better pll values resulting in better
frequencies for the connected devices.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |\ \ |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This patch replaces OK/FAIL by true/false which is simpler and saner.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We do not need viafb_second{,_virtual}_{xres,yres} outside of
viafbdev.c so move them there and eliminate the virtual ones where
the only sane usage is done during initalization.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This patch moves all unprotected VGA initialization in one table and
provides some documentation for those values.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This patch removes the max_hres and max_vres which are not used at
the moment. In general they could be useful but it would be better
to get them via any standard EDID implementation and not the buggy
incomplete one currently used which is also removed as far as
possible.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
As the iga path is the only remaining information which is also
handled by the active devices there is no reason to keep it.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Later the correct values will be written so there is no need to
write early some values which might be wrong.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | | |/
| | |
| | |
| | |
| | |
| | |
| | | |
This moves some mode independend initialization code to the function
where the other parts of the initialization are.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |\ \ |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When allowing some PLL calculation we get a frequency that seems to
be a bit higher than what the OLPC DCON likes resulting in a still
readable but not so good image. We don't really know whether this is
a problem with the calculation formula or the OLPC but as other
displays seem to be happy with the other modes adjusting the OLPC
refresh looks like the better thing. This patch prevents a
regression when dynamic PLL calculation is allowed.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This patch fixes the devices connected on OLPC. The OLPC panel seems
to be connected to DVP1 and LVDS2 for some reasons and if not both
are handled correct the display does not work correct or not at all.
This patch prevents regressions on the OLPC where it worked by
accident but would break in future as the driver did not know the
correct devices connected. This might also fix hardware scaling.
Hopefully the OLPC is the only device with such a requirement but it
will be certainly better to actually know what devices are actually
connected and to not work by accident.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
|
| | |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch fixes a regression introduced by
fd3cc69848b7e1873e5f12bbcdd572b20277ecf3a
"viafb: remove duplicated clock storage"
caused by an incosistent mode which pretended to have a higher
refresh rate than it actually had. The wrong refresh rate resulted
in a calculated higher pixclock which the OLPC DCON could not handle.
By reducing the refresh rate to 50Hz we get close to the old
pixclock which makes the OLPC display usable again.
Minor other adjustments are needed as 60Hz is assumed to be a safe
value which is not true for OLPC DCON. This is no problem as we only
support 1200x900 on the OLPC.
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Reported-by: Daniel Drake <dsd@laptop.org>
|
| |\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6:
[PARISC] wire up syncfs syscall
[PARISC] wire up the fhandle syscalls
[PARISC] wire up clock_adjtime syscall
[PARISC] wire up fanotify syscalls
[PARISC] prevent speculative re-read on cache flush
[PARISC] only make executable areas executable
[PARISC] fix pacache .size with new binutils
|
| | |\ \ |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
According to Appendix F, the TLB is the primary arbiter of speculation.
Thus, if a page has a TLB entry, it may be speculatively read into the
cache. On linux, this can cause us incoherencies because if we're about
to do a disk read, we call get_user_pages() to do the flush/invalidate
in user space, but we still potentially have the user TLB entries, and
the cache could speculate the lines back into userspace (thus causing
stale data to be used). This is fixed by purging the TLB entries before
we flush through the tmpalias space. Now, the only way the line could
be re-speculated is if the user actually tries to touch it (which is not
allowed).
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Currently parisc has the whole kernel marked as RWX, meaning any
kernel page at all is eligible to be executed. This can cause a
theoretical problem on systems with combined I/D TLB because the act
of referencing a page causes a TLB insertion with an executable bit.
This TLB entry may be used by the CPU as the basis for speculating the
page into the I-Cache. If this speculated page is subsequently used
for a user process, there is the possibility we will get a stale
I-cache line picked up as the binary executes.
As a point of good practise, only mark actual kernel text pages as
executable. The same has to be done for init_text pages, but they're
converted to data pages (and the I-Cache flushed) when the init memory
is released.
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Cc: stable@kernel.org
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Cc: stable@kernel.org
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Cc: stable@kernel.org
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Cc: stable@kernel.org
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
|
| | |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fix style of flush_user_dcache_range_asm procedure declaration in
arch/parisc/kernel/pacache.s to be consistent with other assembly
procedures.
Signed-off-by: Meelis Roos <mroos@linux.ee>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes build errors on s390 and probably other archs as well:
In file included from net/ipv4/ip_forward.c:32:0:
include/net/udp.h: In function 'udp_csum_outgoing':
include/net/udp.h:141:2: error: implicit declaration of function 'prefetch'
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes this build error on s390 and probably other archs as well:
fs/inode.c: In function 'new_inode':
fs/inode.c:894:2: error: implicit declaration of function 'spin_lock_prefetch'
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
[ Happens on architectures that don't define their own prefetch
functions in <asm/processor.h>, and instead rely on the default
ones in <linux/prefetch.h> - Linus]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
| |\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
firewire: sbp2: parallelize login, reconnect, logout
firewire: sbp2: octlet AT payloads can be stack-allocated
firewire: sbp2: omit Scsi_Host lock from queuecommand
firewire: core: use non-reentrant workqueue with rescuer
firewire: optimize iso queueing by setting wake only after the last packet
firewire: octlet AT payloads can be stack-allocated
firewire: ohci: optimize find_branch_descriptor()
firewire: ohci: avoid separate DMA mapping for small AT payloads
firewire: ohci: do not start DMA contexts before link is enabled
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The struct sbp2_logical_unit.work items can all be executed in parallel
but are not reentrant. Furthermore, reconnect or re-login work must be
executed in a WQ_MEM_RECLAIM workqueue.
Hence replace the old single-threaded firewire-sbp2 workqueue by a
concurrency-managed but non-reentrant workqueue with rescuer.
firewire-core already maintains one, hence use this one.
In earlier versions of this change, I observed occasional failures of
parallel INQUIRY to an Initio INIC-2430 FireWire 800 to dual IDE bridge.
More testing indicates that parallel INQUIRY is not actually a problem,
but too quick successions of logout and login + INQUIRY, e.g. a quick
sequence of cable plugout and plugin, can result in failed INQUIRY.
This does not seem to be something that should or could be addressed by
serialization.
Another dual-LU device to which I currently have access to, an
OXUF924DSB FireWire 800 to dual SATA bridge with firmware from MacPower,
has been successfully tested with this too.
This change is beneficial to environments with two or more FireWire
storage devices, especially if they are located on the same bus.
Management tasks that should be performed as soon and as quickly as
possible, especially reconnect, are no longer held up by tasks on other
devices that may take a long time, especially login with INQUIRY and sd
or sr driver probe.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We do not need slab allocations for ORB pointer write transactions
anymore in order to satisfy streaming DMA mapping constraints, thanks to
commit da28947e7e36 "firewire: ohci: avoid separate DMA mapping for
small AT payloads".
(Besides, the slab-allocated buffers that firewire-sbp2 used to provide
for 8-byte write requests were still not fully portable since they
shared a cacheline with unrelated CPU-accessed data.)
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
firewire-sbp2 already takes care for internal serialization where
required (ORB list accesses), and it does not use cmd->serial_number
internally. Hence it is safe to not grab the shost lock around
queuecommand.
While we are at housekeeping, drop a redundant struct member:
sbp2_command_orb.done is set once in a hot path and dereferenced once in
a hot path. We can as well dereference sbp2_command_orb.cmd->scsi_done
instead.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
firewire-core manages the following types of work items:
fw_card.br_work:
- resets the bus on a card and possibly sends a PHY packet before that
- does not sleep for long or not at all
- is scheduled via fw_schedule_bus_reset() by
- firewire-ohci's pci_probe method
- firewire-ohci's set_config_rom method, called by kernelspace
protocol drivers and userspace drivers which add/remove
Configuration ROM descriptors
- userspace drivers which use the bus reset ioctl
- itself if the last reset happened less than 2 seconds ago
fw_card.bm_work:
- performs bus management duties
- usually does not (but may in corner cases) sleep for long
- is scheduled via fw_schedule_bm_work() by
- firewire-ohci's self-ID-complete IRQ handler tasklet
- firewire-core's fw_device.work instances whenever the root node
device was (successfully or unsuccessfully) discovered,
refreshed, or rediscovered
- itself in case of resource allocation failures or in order to
obey the 125ms bus manager arbitration interval
fw_device.work:
- performs node probe, update, shutdown, revival, removal; including
kernel driver probe, update, shutdown and bus reset notification to
userspace drivers
- usually sleeps moderately long, in corner cases very long
- is scheduled by
- firewire-ohci's self-ID-complete IRQ handler tasklet via the
core's fw_node_event
- firewire-ohci's pci_remove method via core's fw_destroy_nodes/
fw_node_event
- itself during retries, e.g. while a node is powering up
iso_resource.work:
- accesses registers at the Isochronous Resource Manager node
- usually does not (but may in corner cases) sleep for long
- is scheduled via schedule_iso_resource() by
- the owning userspace driver at addition and removal of the
resource
- firewire-core's fw_device.work instances after bus reset
- itself in case of resource allocation if necessary to obey the
1000ms reallocation period after bus reset
fw_card.br_work instances should not, and instances of the others must
not, be executed in parallel by multiple CPUs -- but were not protected
against that. Hence allocate a non-reentrant workqueue for them.
fw_device.work may be used in the memory reclaim path in case of SBP-2
device updates. Hence we need a workqueue with rescuer and cannot use
system_nrt_wq.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Reviewed-by: Tejun Heo <tj@kernel.org>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When queueing iso packets, the run time is dominated by the two
MMIO accesses that set the DMA context's wake bit. Because most
drivers submit packets in batches, we can save much time by
removing all but the last wakeup.
The internal kernel API is changed to require a call to
fw_iso_context_queue_flush() after a batch of queued packets.
The user space API does not change, so one call to
FW_CDEV_IOC_QUEUE_ISO must specify multiple packets to take
advantage of this optimization.
In my measurements, this patch reduces the time needed to queue
fifty skip packets from userspace to one sixth on a 2.5 GHz CPU,
or to one third at 800 MHz.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We do not need slab allocations anymore in order to satisfy
streaming DMA mapping constraints, thanks to commit da28947e7e36
"firewire: ohci: avoid separate DMA mapping for small AT payloads".
(Besides, the slab-allocated buffers that firewire-core, firewire-sbp2,
and firedtv used to provide for 8-byte write and lock requests were
still not fully portable since they crossed cacheline boundaries or
shared a cacheline with unrelated CPU-accessed data. snd-firewire-lib
got this aspect right by using an extra kmalloc/ kfree just for the
8-byte transaction buffer.)
This change replaces kmalloc'ed lock transaction scratch buffers in
firewire-core, firedtv, and snd-firewire-lib by local stack allocations.
Perhaps the most notable result of the change is simpler locking because
there is no need to serialize usages of preallocated per-device buffers
anymore. Also, allocations and deallocations are simpler.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Acked-by: Clemens Ladisch <clemens@ladisch.de>
|
| | |\ \ \
| | | | |
| | | | |
| | | | |
| | | | | |
in order to pull in changes in drivers/media/dvb/firewire/ and
sound/firewire/.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When z==2, the condition "key == 2" is superfluous because it cannot
occur without "b == 3", as a descriptor with b!=3 and key==2 would be
an OUTPUT_MORE_IMMEDIATE descriptor which cannot be used alone.
Also remove magic numbers and needless computations on the b field.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
For AT packet payloads of up to eight bytes, we have enough unused space
in the DMA descriptors list so that we can put a copy of the payload
there and thus avoid having to create a separate streaming DMA mapping
for the payload buffer.
In a CPU-bound microbenchmark that just sends 8-byte packets, bandwidth
was measured to increase by 5.7 %, from 1009 KB/s to 1067 KB/s. In
practice, the only performance-sensitive usage of small asynchronous
packets is the SBP-2 driver's write to the ORB_POINTER register during
SCSI command submission.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
OHCI 1.1 5.7.3 not only forbids enabling or starting any DMA contexts
before the linkEnable bit is set, but also explicitly warns of undefined
behaviour if this order is violated.
Don't violate it then.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
|
