This converts ide_cd_queue_pc to use blk_execute_rq, necessitating
changing the ide_cd_queue_pc prototype into a form that doesn't takes
a pointer to request struct. ide_cd_queue_pc works like scsi_execute.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Borislav Petkov <petkovbb@gmail.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Call __elv_add_request() with 'plug' == 1 (so the device will be
plugged) and then use __generic_unplug_device() instead of calling
ide_do_request() directly.
v2:
* For blk_pm_resume_request() requests the queue is stopped so we
need to call ->request_fn explicitly.
Thanks to:
- Rafael for reporting/bisecting the bug
- Borislav/Rafael for testing the fix
This is a preparation for converting IDE to use blk_execute_rq().
Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Borislav Petkov <petkovbb@gmail.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* Set REQ_PREEMPT request flag in ide_do_drive_cmd() users
for ide_preempt and ide_head_wait action types.
* Remove setting REQ_PREEMPT from ide_do_drive_cmd().
While at it:
* Set 'where' variable outside ide_lock.
This is a preparation for converting IDE to use blk_execute_rq().
There should be no functional changes caused by this patch.
Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Borislav Petkov <petkovbb@gmail.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Replace the FIT() macro with the kernel-provided clamp_val() macro.
FIT was always being called with a member of a struct ide_timing,
which are shorts, and constant constraints for the min and max.
Thus we can use clamp_val, rather than clamp_t.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Since ide_driveid_update() uses polling to execute the IDENTIFY DEVICE command
but clears nIEN bit in the control register and doesn't mask the IDE interrupt,
the latter does happen and lead to the corresponding message to appear:
ide0: unexpected interrupt, status=0x58, count=1
when e.g. running hdparm with option -X with a non-PCI IDE driver...
Signed-off-by: Mikhail Cherkashin <mcherkashin@ru.mvista.com>
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Fix tRP timing for UDMA mode 4 according to the ATA specification.
Signed-off-by: Mikhail Cherkashin <mcherkhashin@ru.mvista.com>
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* 'timers/for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86: add PCI ID for 6300ESB force hpet
x86: add another PCI ID for ICH6 force-hpet
kernel-paramaters: document pmtmr= command line option
acpi_pm clccksource: fix printk format warning
nohz: don't stop idle tick if softirqs are pending.
pmtmr: allow command line override of ioport
nohz: reduce jiffies polling overhead
hrtimer: Remove unused variables in ktime_divns()
hrtimer: remove warning in hres_timers_resume
posix-timers: print RT watchdog message
* 'core/topology' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
cputopology: always define CPU topology information, clean up
cpu topology: always define CPU topology information
Do not automatically "select" SCSI_DH for dm-multipath. If SCSI_DH
doesn't exist,just do not allow hardware handlers to be used.
Handle SCSI_DH being a module also. Make sure it doesn't allow DM_MULTIPATH
to be compiled in when SCSI_DH is a module.
[jejb: added comment for Kconfig syntax]
Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Current code uses kmalloc() and then just does a bitwise OR operation on
qp->flags in create_qp_common(), which means that qp->flags may
potentially have some unintended bits set. This patch uses kzalloc()
and avoids further explicit clearing of structure members, which also
shrinks the code:
add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-65 (-65)
function old new delta
create_qp_common 2024 1959 -65
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
There was a bug in some versions of the mlx4 driver in
mlx4_alloc_fmr(), which hardcoded the minimum acceptable page_shift to
be 12. However, new ConnectX firmware can support a minimum
page_shift of 9 (log_pg_sz of 9 returned by QUERY_DEV_LIM) -- so with
old drivers, ib_fmr_alloc() would fail for ULPs using the device
minimum when creating FMRs.
To preserve firmware compatibility with released mlx4 drivers, the
firmware will continue to return 12 as before for log_page_sz in
QUERY_DEV_CAP for these drivers. However, to enable new drivers to
take advantage of the available smaller page size, the mlx4 driver now
first sets the log_pg_sz to the device minimum by setting a
log_page_sz value to 0 via the MOD_STAT_CFG command and then reading
the real minimum via QUERY_DEV_CAP.
Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il>
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The RDMA CM has some logic in place to make sure that callbacks on a
given CM ID are delivered to the consumer in a serialized manner.
Specifically it has code to protect against a device removal racing
with a running callback function.
This patch simplifies this logic by using a mutex per ID instead of a
wait queue and atomic variable. This means that cma_disable_remove()
now is more properly named to cma_disable_callback(), and
cma_enable_remove() can now be removed because it just would become a
trivial wrapper around mutex_unlock().
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Keep a pointer to the local (src) netdevice in struct rdma_dev_addr,
and copy it in as part of rdma_copy_addr(). Use rdma_translate_ip()
in cma_new_conn_id() to reduce some code duplication and also make
sure the src_dev member gets set.
In a high-availability configuration the netdevice pointer can be used
by the RDMA CM to align RDMA sessions to use the same links as the IP
stack does under fail-over and route change cases.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Handling the zero STag in receive work request requires some extra
logic in the driver:
- Only set the QP_PRIV bit for kernel mode QPs.
- Add a zero STag build function for recv wrs. The uP needs a PBL
allocated and passed down in the recv WR so it can construct a HW
PBL for the zero STag S/G entries. Note: we need to place a few
restrictions on zero STag usage because of this:
1) all SGEs in a recv WR must either be zero STag or not. No mixing.
2) an individual SGE length cannot exceed 128MB for a zero-stag SGE.
This should be OK since it's not really practical to allocate
such a large chunk of pinned contiguous DMA mapped memory.
- Add an optimized non-zero-STag recv wr format for kernel users.
This is needed to optimize both zero and non-zero STag cracking in
the recv path for kernel users.
- Remove the iwch_ prefix from the static build functions.
- Bump required FW version.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
- Change the IB_DEVICE_ZERO_STAG flag to the transport-neutral name
IB_DEVICE_LOCAL_DMA_LKEY, which is used by iWARP RNICs to indicate 0
STag support and IB HCAs to indicate reserved L_Key support.
- Add a u32 local_dma_lkey member to struct ib_device. Drivers fill
this in with the appropriate local DMA L_Key (if they support it).
- Fix up the drivers using this flag.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The MLX transport requires two extra gather entries for sends (one for
the header and one for the checksum at the end, as the comment says).
However the code checked that max_recv_sge was not too big, instead of
checking max_send_sge as it should have. Fix the code to check the
correct condition.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Exactly when the catastrophic error polling timer function runs is not
important, so use round_jiffies() to save unnecessary wakeups.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since we use del_timer_sync() anyway, there's no need for an
additional flag to tell the timer not to rearm.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Increase IPoIB ring sizes to twice their original sizes (RX: 128->256,
TX: 64->128) to act as a shock absorber for high traffic peaks. With
the current settings, we have seen cases that there are many calls to
netif_stop_queue(), which causes degradation in throughput. Also,
larger receive buffer sizes help IPoIB in CM mode to avoid experiencing
RNR NAK conditions due to insufficient receive buffers at the SRQ.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since IPoIB connected mode does not NETIF_F_SG, we only have one DMA
mapping per send, so we don't need a mapping[] array. Define a new
struct with a single u64 mapping member and use it for the CM tx_ring.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IB spe. for SubnGet(NodeInfo) and query HCA says that the vendor
ID field should be the IEEE OUI assigned to the vendor. The ipath
driver was returning the PCI vendor ID instead. This will affect
applications which call ibv_query_device(). The old value was
0x001fc1 or 0x001077, the new value is 0x001175.
The vendor ID doesn't appear to be exported via /sys so that should
reduce possible compatibility issues. I'm only aware of Open MPI as a
major application which depends on this change, and they have made
necessary adjustments.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the driver sets the MTU of the net device outside of its
change_mtu method, it should make use of dev_set_mtu() instead of
directly setting the mtu field of struct netdevice. Otherwise
functions registered to be called upon MTU change will not get called
(this is done through call_netdevice_notifiers() in dev_set_mtu()).
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use of this lock is required to synchronize changes to the netdvice's
data structs. Also move the call to ipoib_flush_paths() after the
modification of the netdevice flags in set_mode().
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipoib_mcast_detach() does nothing except call ib_detach_mcast(), so just
use the core API in the one place that does a multicast group detach.
add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-105 (-105)
function old new delta
ipoib_mcast_leave 357 319 -38
ipoib_mcast_detach 67 - -67
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The current code will set the Q_Key for any join of a non-sendonly
multicast group. The operation involves a modify QP operation, which
is fairly heavyweight, and is only really required after the join of
the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach()
to control when the Q_Key is set.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
No need for a mutex around calls to ib_attach_mcast/ib_detach_mcast
since these operations are synchronized at the HW driver layer.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IPOIB_MCAST_STARTED flag is not used at all since commit b3e2749b
("IPoIB: Don't drop multicast sends when they can be queued"), so
remove it.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Every caller of nes_post_cqp_request() passed it NES_CQP_REQUEST_RING_DOORBELL,
so just remove that parameter and always ring the doorbell.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Faisal Latif <flatif@neteffect.com>
cxgb3 does not currently report the page size capabilities, and
incorrectly reports them internally.
This version changes the bit-shifting to a static value (per Steve's
request).
Signed-off-by: Jon Mason <jon@opengridcomputing.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The patch tries to solve the problem of device going down and paths being
flushed on an SM change event. The method is to mark the paths as candidates for
refresh (by setting the new valid flag to 0), and wait for an ARP
probe a new path record query.
The solution requires a different and less intrusive handling of SM
change event. For that, the second argument of the flush function
changes its meaning from a boolean flag to a level. In most cases, SM
failover doesn't cause LID change so traffic won't stop. In the rare
cases of LID change, the remote host (the one that hadn't changed its
LID) will lose connectivity until paths are refreshed. This is no
worse than the current state. In fact, preventing the device from
going down saves packets that otherwise would be lost.
Signed-off-by: Moni Levy <monil@voltaire.com>
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This gives ehca an autogenerated modalias and therefore enables automatic loading.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add "ipoib_use_lro" module parameter to enable LRO and an
"ipoib_lro_max_aggr" module parameter to set the max number of packets
to be aggregated. Make LRO controllable and LRO statistics accessible
through ethtool.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Set IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK for IPoIB's UD QPs if
supported by the underlying device. This creates an improvement of up
to 39% in bandwidth when sending multicast packets with IPoIB, and an
improvment of 12% in cpu usage.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add support for handling the IB_QP_CREATE_MULTICAST_BLOCK_LOOPBACK
flag by using the per-multicast group loopback blocking feature of
mlx4 hardware.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- Add a new rdma ctl command called RDMA_GET_MIB to the cxgb3 low
level driver to obtain the protocol mib from the rnic hardware.
- Add new iw_cxgb3 provider method to get the MIB from the low level
driver.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds a sysfs attribute group called "proto_stats" under
/sys/class/infiniband/$device/ and populates this group with protocol
statistics if they exist for a given device. Currently, only iWARP
stats are defined, but the code is designed to allow InfiniBand
protocol stats if they become available. These stats are per-device
and more importantly -not- per port.
Details:
- Add union rdma_protocol_stats in ib_verbs.h. This union allows
defining transport-specific stats. Currently only iwarp stats are
defined.
- Add struct iw_protocol_stats to define the current set of iwarp
protocol stats.
- Add new ib_device method called get_proto_stats() to return protocol
statistics.
- Add logic in core/sysfs.c to create iwarp protocol stats attributes
if the device is an RNIC and has a get_proto_stats() method.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
For devices that don't support SRQs, ipoib_cm_post_receive_nonsrq() is
called from both ipoib_cm_handle_rx_wc() and ipoib_cm_nonsrq_init_rx(),
and these two callers are not synchronized against each other.
However, ipoib_cm_post_receive_nonsrq() always reuses the same receive
work request and scatter list structures, so multiple callers can end
up stepping on each other, which leads to posting garbled work
requests.
Fix this by having the caller pass in the ib_recv_wr and ib_sge
structures to use, and allocating new local structures in
ipoib_cm_nonsrq_init_rx().
Based on a patch by Pradeep Satyanarayana <pradeep@us.ibm.com> and
David Wilder <dwilder@us.ibm.com>, with debugging help from Hoang-Nam
Nguyen <hnguyen@de.ibm.com>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The members struct iwch_rnic_attributes.vendor_id and .vendor_part_id
are write-only, so we might as well get rid of them.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
- set fw_ver
- set hw_ver
- set max_qp_wr to something reasonable
- set max_cqe to something reasonable
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
During corner case testing, we noticed that some versions of ehca do
not properly transition to interrupt done in special load situations.
This can be resolved by periodically triggering EOI through H_EOI, if
EQEs are pending.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit 65adfa91 ("IB/mlx4: Fix RESET to RESET and RESET to ERROR
transitions") added some extra code to handle a QP state transition
from RESET to ERROR. However, the latest 1.2.1 version of the IB spec
has clarified that this transition is actually not allowed, so we can
remove this extra code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit b18aad71 ("IB/mthca: Fix RESET to ERROR transition") added some
extra code to handle a QP state transition from RESET to ERROR.
However, the latest 1.2.1 version of the IB spec has clarified that
this transition is actually not allowed, so we can remove this extra
code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
I was reviewing the QP state transition diagram in the IB 1.2.1 spec
and the code for qp_state_table[], and noticed that the code allows a
QP to be modified from IB_QPS_RESET to IB_QPS_ERR whereas the notes
for figure 124 (pg 457) specifically says that this transition isn't
allowed. This is a clarification from earlier versions of the IB
spec, which were ambiguous in this area and suggested that the RESET
to ERR transition was allowed.
Fix up the qp_state_table[] to make RESET->ERR not allowed.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX HCAs support the IB_MGMT_CLASS_CONG_MGMT management class, so
process MADs of this class through the MAD_IFC firmware command.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX returns the max message size it supports through the
QUERY_DEV_CAP firmware command. When modifying a QP to RTR, the max
message size for the QP must be specified. This value must not exceed
the value declared through QUERY_DEV_CAP. The current code ignores
the max allowed size and unconditionally sets the value to 2^31. This
patch sets all QPs to the max value allowed as returned from firmware.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- set IB_DEVICE_MEM_MGT_EXTENSIONS capability bit if fw supports it.
- set max_fast_reg_page_list_len device attribute.
- add iwch_alloc_fast_reg_mr function.
- add iwch_alloc_fastreg_pbl
- add iwch_free_fastreg_pbl
- adjust the WQ depth for kernel mode work queues to account for
fastreg possibly taking 2 WR slots.
- add fastreg_mr work request support.
- add local_inv work request support.
- add send_with_inv and send_with_se_inv work request support.
- removed useless duplicate enums/defines for TPT/MW/MR stuff.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds support for the IB "base memory management extension"
(BMME) and the equivalent iWARP operations (which the iWARP verbs
mandates all devices must implement). The new operations are:
- Allocate an ib_mr for use in fast register work requests.
- Allocate/free a physical buffer lists for use in fast register work
requests. This allows device drivers to allocate this memory as
needed for use in posting send requests (eg via dma_alloc_coherent).
- New send queue work requests:
* send with remote invalidate
* fast register memory region
* local invalidate memory region
* RDMA read with invalidate local memory region (iWARP only)
Consumer interface details:
- A new device capability flag IB_DEVICE_MEM_MGT_EXTENSIONS is added
to indicate device support for these features.
- New send work request opcodes IB_WR_FAST_REG_MR, IB_WR_LOCAL_INV,
IB_WR_RDMA_READ_WITH_INV are added.
- A new consumer API function, ib_alloc_mr() is added to allocate
fast register memory regions.
- New consumer API functions, ib_alloc_fast_reg_page_list() and
ib_free_fast_reg_page_list() are added to allocate and free
device-specific memory for fast registration page lists.
- A new consumer API function, ib_update_fast_reg_key(), is added to
allow the key portion of the R_Key and L_Key of a fast registration
MR to be updated. Consumers call this if desired before posting
a IB_WR_FAST_REG_MR work request.
Consumers can use this as follows:
- MR is allocated with ib_alloc_mr().
- Page list memory is allocated with ib_alloc_fast_reg_page_list().
- MR R_Key/L_Key "key" field is updated with ib_update_fast_reg_key().
- MR made VALID and bound to a specific page list via
ib_post_send(IB_WR_FAST_REG_MR)
- MR made INVALID via ib_post_send(IB_WR_LOCAL_INV),
ib_post_send(IB_WR_RDMA_READ_WITH_INV) or an incoming send with
invalidate operation.
- MR is deallocated with ib_dereg_mr()
- page lists dealloced via ib_free_fast_reg_page_list().
Applications can allocate a fast register MR once, and then can
repeatedly bind the MR to different physical block lists (PBLs) via
posting work requests to a send queue (SQ). For each outstanding
MR-to-PBL binding in the SQ pipe, a fast_reg_page_list needs to be
allocated (the fast_reg_page_list is owned by the low-level driver
from the consumer posting a work request until the request completes).
Thus pipelining can be achieved while still allowing device-specific
page_list processing.
The 32-bit fast register memory key/STag is composed of a 24-bit index
and an 8-bit key. The application can change the key each time it
fast registers thus allowing more control over the peer's use of the
key/STag (ie it can effectively be changed each time the rkey is
rebound to a page list).
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The connected mode implementation in the IPoIB driver has a large
overhead in the way SKBs are handled in the receive flow. It usually
allocates an SKB with as big as was used in the currently received SKB
and moves unused fragments from the old SKB to the new one. This
involves a loop on all the remaining fragments and incurs overhead on
the CPU. This patch, for small SKBs, allocates an SKB just large
enough to contain the received data and copies to it the data from the
received SKB. The newly allocated SKB is passed to the stack and the
old SKB is reposted.
When running netperf, UDP small messages, without this pach I get:
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
14.4.3.178 (14.4.3.178) port 0 AF_INET
Socket Message Elapsed Messages
Size Size Time Okay Errors Throughput
bytes bytes secs # # 10^6bits/sec
114688 128 10.00 5142034 0 526.31
114688 10.00 1130489 115.71
With this patch I get both send and receive at ~315 mbps.
The reason that send performance actually slows down is as follows:
When using this patch, the overhead of the CPU for handling RX packets
is dramatically reduced. As a result, we do not experience RNR NAK
messages from the receiver which cause the connection to be closed and
reopened again; when the patch is not used, the receiver cannot handle
the packets fast enough so there is less time to post new buffers and
hence the mentioned RNR NACKs. So what happens is that the
application *thinks* it posted a certain number of packets for
transmission but these packets are flushed and do not really get
transmitted. Since the connection gets opened and closed many times,
each time netperf gets the CPU time that otherwise would have been
given to IPoIB to actually transmit the packets. This can be verified
when looking at the port counters -- the output of ifconfig and the
oputput of netperf (this is for the case without the patch):
tx packets
==========
port counter: 1,543,996
ifconfig: 1,581,426
netperf: 5,142,034
rx packets
==========
netperf 1,1304,089
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
The idea is that for QPs with fixed size work requests (eg selective
signaling QPs), before stamping the WQE, we read the value of the DS
field, which gives the effective size of the descriptor as used in the
previous post. Then we stamp only that area, since the rest of the
descriptor is already stamped.
When initializing the send queue buffer, make sure the DS field is
initialized to the max descriptor size so that the subsequent stamping
will be done on the entire descriptor area.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch solves a race that occurs after an event occurs that causes
the SA query module to flush its SM address handle (AH). When SM AH
becomes invalid and needs an update it is handled by the global
workqueue. On the other hand this event is also handled in the IPoIB
driver by queuing work in the ipoib_workqueue that does multicast
joins. Although queuing is in the right order, it is done to 2
different workqueues and so there is no guarantee that the first to be
queued is the first to be executed.
This causes a problem because IPoIB may end up sending an request to
the old SM, which will take a long time to time out (since the old SM
is gone); this leads to a much longer than necessary interruption in
multicast traffer.
The patch sets the SA query module's SM AH to NULL when the event
occurs, and until update_sm_ah() is done, any request that needs sm_ah
fails with -EAGAIN return status.
For consumers, the patch doesn't make things worse. Before the patch,
MADs are sent to the wrong SM so the request gets lost. Consumers can
be improved if they examine the return code and respond to EAGAIN
properly but even without an improvement the situation is not getting
worse.
Signed-off-by: Moni Levy <monil@voltaire.com>
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The license text for several files references a third software license
that was inadvertently copied in. Update the license to what was
intended. This update was based on a request from HP.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove an explicit memset(..., 0, ...) of a 'listener' structure
allocated with kzalloc().
Signed-off-by: Christophe Jaillet <christophe.jaillet@wanadoo.fr>
Acked-by: Faisal Latif <faisal@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The SRP initiator is currently using ib_find_cached_pkey() and
ib_get_cached_gid() in situations where the uncached ib_find_pkey()
and ib_query_gid() functions serve just as well: sleeping is allowed
and performance is not an issue. Since we want to eliminate the
cached operations in the long term, convert SRP to use the uncached
variants.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
With new userspace libpciaccess we can get a conflicting mapping
on the PCIE GART table in the video RAM. Always try and map it _wc.
Signed-off-by: Dave Airlie <airlied@redhat.com>
This patch adds support to the RGMII handler in the EMAC driver for
the MII PHY mode such that device tree entries of the form `phy-mode = "mii";'
are recognized and handled appropriately.
While logically, in software, "gmii" and "mii" modes are the same,
they are wired differently, so it makes sense to allow DTS authors to
specify each explicitly.
Signed-off-by: Grant Erickson <gerickson@nuovations.com>
Acked-by: Stefan Roese <sr@denx.de>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Adds a character driver for BSR support on IBM POWER systems including
Power5 and Power6. The BSR is an optional processor facility not currently
implemented by any other processors. It's primary purpose is fast large SMP
synchronization. More details on the BSR are in comments to the code which
follows. This patch adds BSR driver to pseries_defconfig.
Signed-off-by: Sonny Rao <sonnyrao@linux.vnet.ibm.com>
Signed-off-by: Joel Schopp <jschopp@austin.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
* 'for-2.6.27' of git://git.infradead.org/users/dwmw2/firmware-2.6: (64 commits)
firmware: convert sb16_csp driver to use firmware loader exclusively
dsp56k: use request_firmware
edgeport-ti: use request_firmware()
edgeport: use request_firmware()
vicam: use request_firmware()
dabusb: use request_firmware()
cpia2: use request_firmware()
ip2: use request_firmware()
firmware: convert Ambassador ATM driver to request_firmware()
whiteheat: use request_firmware()
ti_usb_3410_5052: use request_firmware()
emi62: use request_firmware()
emi26: use request_firmware()
keyspan_pda: use request_firmware()
keyspan: use request_firmware()
ttusb-budget: use request_firmware()
kaweth: use request_firmware()
smctr: use request_firmware()
firmware: convert ymfpci driver to use firmware loader exclusively
firmware: convert maestro3 driver to use firmware loader exclusively
...
Fix up trivial conflicts with BKL removal in drivers/char/dsp56k.c and
drivers/char/ip2/ip2main.c manually.
* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm: (241 commits)
[ARM] 5171/1: ep93xx: fix compilation of modules using clocks
[ARM] 5133/2: at91sam9g20 defconfig file
[ARM] 5130/4: Support for the at91sam9g20
[ARM] 5160/1: IOP3XX: gpio/gpiolib support
[ARM] at91: Fix NAND FLASH timings for at91sam9x evaluation kits.
[ARM] 5084/1: zylonite: Register AC97 device
[ARM] 5085/2: PXA: Move AC97 over to the new central device declaration model
[ARM] 5120/1: pxa: correct platform driver names for PXA25x and PXA27x UDC drivers
[ARM] 5147/1: pxaficp_ir: drop pxa_gpio_mode calls, as pin setting
[ARM] 5145/1: PXA2xx: provide api to control IrDA pins state
[ARM] 5144/1: pxaficp_ir: cleanup includes
[ARM] pxa: remove pxa_set_cken()
[ARM] pxa: allow clk aliases
[ARM] Feroceon: don't disable BPU on boot
[ARM] Orion: LED support for HP mv2120
[ARM] Orion: add RD88F5181L-FXO support
[ARM] Orion: add RD88F5181L-GE support
[ARM] Orion: add Netgear WNR854T support
[ARM] s3c2410_defconfig: update for current build
[ARM] Acer n30: Minor style and indentation fixes.
...
This includes PXA work up to the SPI changes for the initial merge,
since e172274ccc depends on the SPI
tree being merged.
Conflicts:
arch/arm/configs/em_x270_defconfig
arch/arm/configs/xm_x270_defconfig
Version number provided in first HEX record.
Signed-off-by: Jaswinder Singh <jaswinder@infradead.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Although it wasn't actually using ihex records before, we use the Intel
HEX record format for this firmware -- because that gives us a simple
way to split it into separate chunks internally as we need, without
loading each part as a separate file.
Signed-off-by: Jaswinder Singh <jaswinder@infradead.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Add a mechanism to let new-style i2c drivers optionally autodetect
devices they would support on selected buses and ask i2c-core to
instantiate them. This is a replacement for legacy i2c drivers, much
cleaner.
Where drivers had to implement both a legacy i2c_driver and a
new-style i2c_driver so far, this mechanism makes it possible to get
rid of the legacy i2c_driver and implement both enumerated and
detected device support with just one (new-style) i2c_driver.
Here is a quick conversion guide for these drivers, step by step:
* Delete the legacy driver definition, registration and removal.
Delete the attach_adapter and detach_client methods of the legacy
driver.
* Change the prototype of the legacy detect function from
static int foo_detect(struct i2c_adapter *adapter, int address, int kind);
to
static int foo_detect(struct i2c_client *client, int kind,
struct i2c_board_info *info);
* Set the new-style driver detect callback to this new function, and
set its address_data to &addr_data (addr_data is generally provided
by I2C_CLIENT_INSMOD.)
* Add the appropriate class to the new-style driver. This is
typically the class the legacy attach_adapter method was checking
for. Class checking is now mandatory (done by i2c-core.) See
<linux/i2c.h> for the list of available classes.
* Remove the i2c_client allocation and freeing from the detect
function. A pre-allocated client is now handed to you by i2c-core,
and is freed automatically.
* Make the detect function fill the type field of the i2c_board_info
structure it was passed as a parameter, and return 0, on success. If
the detection fails, return -ENODEV.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
We call adapter->client_register for both legacy and new-style i2c
devices, however we only call adapter->client_unregister for legacy
drivers. This doesn't make much sense. Usually, drivers will undo
in client_unregister what they did in client_register, so we should
call neither or both for every given i2c device.
In order to ease the transition from legacy to new-style devices, it
seems preferable to actually call both.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: David Brownell <david-b@pacbell.net>
Clean up old i2c chip drivers:
* Name the i2c_client "client" instead of "new_client".
* Drop useless initializations to 0.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
This patch completes the conversion of the IBM IIC driver to an
of-platform driver.
It removes the index from the IBM IIC driver and makes it an unnumbered
driver. It then calls of_register_i2c_devices to properly register all
the child nodes in the DTS.
Signed-off-by: Sean MacLennan <smaclennan@pikatech.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Add a new-style driver for most I2C EEPROMs, giving sysfs read/write
access to their data. Tested with various chips and clock rates.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Export the root of the i2c bus so that PowerPC device tree code can
iterate over devices on the i2c bus.
Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Fix driver power management:
- suspend the PSC while driver is idle.
- move PSC init/deinit to separate functions, as PSC must be
initialized/shutdown on resume/suspend.
Signed-off-by: Manuel Lauss <mano@roarinelk.homelinux.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Implementing detach_client is optional, so there is no point in
an empty implementation.
Likewise, i2c driver IDs are optional, and we don't need one.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Check for ACPI resource conflicts in i2c bus drivers. I've included
all recent SMBus master drivers for PC hardware.
I've voluntarily left out:
* Drivers that don't run on PCs: they can't conflict with ACPI.
* Bit-banged bus device drivers: it's very unlikely that ACPI would
deal with such buses.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Basic PM support: reinit the core on resume, disable it on suspend.
Signed-off-by: Manuel Lauss <mano@roarinelk.homelinux.net>
Acked-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
The standard rtc-m41t80.c driver cannot be used with the SWARM as it is,
because the board does not provide setup information for the I2C core.
As a result the bus and the address to probe for the M41T80 chip is not
known.
Here is a set of changes that fix the problem:
1. swarm-i2c.c -- SWARM I2C board setup, currently for the M41T80 chip on
the bus #1 only (there is a MAX6654 temperature sensor on the bus #0
which may be added in the future if we have a driver for that chip).
2. The i2c-sibyte.c BCM1250A SMBus controller driver now registers its
buses as numbered so that board setup is correctly applied.
Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Move the check of pre-transaction and post-transaction conditions to
separate functions, and adjust them a bit. Having dedicated functions
for that ensures that errors are handled in a consistent way.
Bit HOST_BUSY of the status register is read-only, so writing to it is
certainly not going to clear it. If this bit is set then we simply
don't want to start the transaction, as it means that somebody else
(ACPI, SMM?) is already using the controller.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Bit BUS_ERR of the status register means that the ICH host controller
lost the arbitration. Report this event as such.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Dumping the register values before and after every transaction was
useful during driver development but now it's only spamming the log.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Struct members udelay and timeout aren't used anywhere, so drop them.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Eric Brower <ebrower@gmail.com>
Improve lost-arbitration handling of PCF8584. This is necessary for
support of a currently out-of-kernel driver for Sun Microsystems E250
environmental management; perhaps others.
Signed-off-by: Eric Brower <ebrower@gmail.com>
Acked-by: Dan Smolik <marvin@mydatex.cz>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
The legacy pcf8574 and pcf8575 drivers should be avoided
on systems using the new gpiolib code, and generally deprecated
in the same way the legacy pca9539 driver is deprecated.
Also, correct the pca9539 deprecation to match the current name
of the preferred driver: pca953x, supporting several more chips.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Initialize the pxa i2c bus during subsystem initialization to make it
available during driver initialization (e.g. display powerup for pxafb).
Signed-off-by: Uli Luckas <u.luckas@road.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Enable the IBM I2C driver for all PPC4xx variants by adding
"ibm,iic" to the compatible list. This way all currently available
arch/powerpc 4xx ports can make use of this driver without any changes.
Additionally all "other" compatible entries are removed since they are
not needed anymore.
Currently all 4xx PPC's have the same compatible I2C macro. If at some
time an incompatibility is detected we can take care of this with an
additional property.
Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
The deprecated OCP style driver part is used by the "old" arch/ppc
platform. This platform is scheduled for removal in June/July this year.
This patch now removes the OCP driver part from the IBM I2C driver.
Signed-off-by: Stefan Roese <sr@denx.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
When I2C block reads are not supported by the underlying adapter, use
SMBus read word transactions instead of consecutive byte reads.
Reasons for this change are:
* The consecutive byte read approach is not safe on multi-master buses.
* While consecutive byte reads have less overhead if you only count the
bytes on the bus, it takes more than twice as many transactions as
with SMBus read word transactions, and each transaction has a cost:
taking and releasing the adapter mutex, and for polling drivers,
waiting for the transaction to complete.
This change yields a significant performance boost at HZ=250 with
EEPROMs on an Intel 82801 bus (basically twice as fast.)
SMBus read word transactions are widely supported so I don't expect
compatibility issues.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Let general purpose I2C/SMBus bus drivers add SPD to their class. Once
this is done, we will be able to tell the eeprom driver to only probe
for SPD EEPROMs and similar on these buses.
Note that I took a conservative approach here, adding I2C_CLASS_SPD to
many drivers that have no idea whether they can host SPD EEPROMs or not.
This is to make sure that the eeprom driver doesn't stop probing buses
where SPD EEPROMs or equivalent live.
So, bus driver maintainers and users should feel free to remove the SPD
class from drivers those buses never have SPD EEPROMs or they don't
want the eeprom driver to bind to them. Likewise, feel free to add the
SPD class to any bus driver I might have missed.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Let framebuffer drivers set their I2C bus class to DDC. Once this is
done, we will be able to tell the eeprom driver to only probe for
EDID EEPROMs on these buses.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Bugfixes to the i2c-cpm driver
- enable correct interrupts (I2CER_TXE instead of I2CER_BUSY)
- replace forgotten iic with i2c
- prefix debug-output on init with 0x and add frequency
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Jochen Friedrich <jochen@scram.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
This driver uses the port of 2.4 code from Vitaly Bordug
<vitb@kernel.crashing.org> and the actual algorithm used by the i2c
driver of the DBox code on cvs.tuxboc.org from Felix Domke
(tmbinc@gmx.net) and Gillem (htoa@gmx.net) converted to an
of_platform_driver. Tested on CPM1 (MPC823 on dbox2 hardware) and
CPM2 (MPC8272).
Signed-off-by: Jochen Friedrich <jochen@scram.de>
Tested-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
This is part of the effort to get rid of the BKL.
[JD: In fact i2c-dev doesn't need more locking than is already done
for the other i2c drivers, so we can simply switch to unlocked_ioctl.]
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
New i2c bus driver for the Intel SCH chipsets (AF82US15W, AF82US15L,
AF82UL11L).
Signed-off-by: Alek Du <alek.du@intel.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Use list_for_each_entry_safe() in i2c_del_adapter() and i2c_del_driver().
Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Fix errorcode to be more descriptive when ioremap fails.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Give a more concrete error code, when the bus is not idle.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Many PC SMBus host controller drivers don't properly handle the case
where they are requested to achieve a transaction they do not support.
Update them so that the consistently print a warning message and
return a single error value in this case.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
The i2c-piix4 driver was used recently as a model to write a new SMBus
host controller driver and this made me realize that the code of this
old driver wasn't exactly good. So, here are many cleanups and minor
fixes to this driver, so that these minor mistakes aren't duplicated
again:
* Delete unused structure.
* Delete needless forward function declaration.
* Properly announce the SMBus host controller as we find it.
* Spell it SMBus not SMB.
* Return -EBUSY instead of -ENODEV when the I/O region is already in
use.
* Drop useless masks on the 7-bit address and the R/W bit.
* Reject block transaction requests with an invalid block length.
* Check and report block transaction replies with an invalid block
length.
* Delete a useless comment.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Tighten error paths used by various i2c adapters (mostly x86) so
they return real fault/errno codes instead of a "-1" (which is
most often interpreted as "-EPERM"). Build tested, with eyeball
review.
One minor initial goal is to have adapters consistently return
the code "-ENXIO" when addressing a device doesn't get an ACK
response, at least in the probe paths where they are already
good at stifling related logspam.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Defend the i2c refcount calls against NULL pointers, as is important
(and conventional) for such calls. Note that none of the current
callers of i2c_use_client() use its return value.
[JD: I hate this but apparently all the other subsystems do it so...]
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Provide kerneldoc for most of the I2C and SMBus I/O calls. Add a
comment summarizing some fault reporting issues which affect the
ability to provide clean fault reports through I2C master transfer
calls. (Making it hard to precisely specify their return values...)
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Function i2c_smbus_write_quick has no users left, so we can delete it.
Also update the list of these helper functions which are gone but
could be added back if needed.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
The list of I2C/SMBus bus drivers is growing and it is sometimes
difficult for the users to figure out what drivers they should enable.
By grouping the drivers by type, I hope to make the selection easier.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
More updates to the I2C stack's fault reporting: make the core stop
returning "-1" (usually "-EPERM") for all faults. Instead, pass lower
level fault code up the stack, or return some appropriate errno.
This patch happens to touch almost exclusively SMBus calls.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Since only a few Blackfins lack TWI, just list them in a depends
statement.
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
This patch contains the scheduled removal of i2c-i810, i2c-prosavage
and i2c-savage4.
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Just like the Tyan S4882, the S4985 uses a multiplexer to give access
to all 16 memory module SPD EEPROMs. This specific i2c-nforce2-s4985
driver adds support for this. It is heavily based on the older
i2c-amd756-s4882 driver.
As more mainboards will use multiplexer chips, we will have to find a
way to support them without having to write a new specfic driver for
each. The recent changes to the i2c subsystem should help us, and the
new gpio subsystem might help, too.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
If an interrupt happens before an I2c master read/write,
complete is called on uninitialized structure.
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
If wait_for_completion_interruptible_timeout exits due
to a signal, the i2c bus was locking up.
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Interrupts are enabled at the point where the DAVINCI_I2C_IVR_REG is read,
so unless an interrupt happened just at that moment, no interrupt would be
pending. Even though documentation implies you should do this, I see no
reason. If slave support is added, this read would cause a hard to
reproduce bug.
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Previously the dev_dbg only printed if no error.
Printing also on an error is more useful
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Ensure psc value gives a clock between 7-12 MHz
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Kevin Hilman <khilman@mvista.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/hskinnemoen/avr32-2.6: (31 commits)
avr32: Fix typo of IFSR in a comment in the PIO header file
avr32: Power Management support ("standby" and "mem" modes)
avr32: Add system device for the internal interrupt controller (intc)
avr32: Add simple SRAM allocator
avr32: Enable SDRAMC clock at startup
rtc-at32ap700x: Enable wakeup
macb: Basic suspend/resume support
atmel_serial: Drain console TX shifter before suspending
atmel_serial: Fix build on avr32 with CONFIG_PM enabled
avr32: Use a quicklist for PTE allocation as well
avr32: Use a quicklist for PGD allocation
avr32: Cover the kernel page tables in the user PGDs
avr32: Store virtual addresses in the PGD
avr32: Remove useless zeroing of swapper_pg_dir at startup
avr32: Clean up and optimize the TLB operations
avr32: Rename at32ap.c -> pdc.c
avr32: Move setup_platform() into chip-specific file
avr32: Kill special exception handler sections
avr32: Kill unneeded #include <asm/pgalloc.h> from asm/mmu_context.h
avr32: Clean up time.c #includes
...
* 'for-linus' of git://git.kernel.dk/linux-2.6-block: (37 commits)
splice: fix generic_file_splice_read() race with page invalidation
ramfs: enable splice write
drivers/block/pktcdvd.c: avoid useless memset
cdrom: revert commit 22a9189 (cdrom: use kmalloced buffers instead of buffers on stack)
scsi: sr avoids useless buffer allocation
block: blk_rq_map_kern uses the bounce buffers for stack buffers
block: add blk_queue_update_dma_pad
DAC960: push down BKL
pktcdvd: push BKL down into driver
paride: push ioctl down into driver
block: use get_unaligned_* helpers
block: extend queue_flag bitops
block: request_module(): use format string
Add bvec_merge_data to handle stacked devices and ->merge_bvec()
block: integrity flags can't use bit ops on unsigned short
cmdfilter: extend default read filter
sg: fix odd style (extra parenthesis) introduced by cmd filter patch
block: add bounce support to blk_rq_map_user_iov
cfq-iosched: get rid of enable_idle being unused warning
allow userspace to modify scsi command filter on per device basis
...
In an I/O heavy workload (IOZone), ahci_qc_issue is the second-highest
consumer of CPU cycles. Removing the flush gets us approximately 10%
bandwidth improvement. I believe this to be because the CPU can start
queueing the next request instead of waiting for the readl() to flush the
writes to the device. The flush isn't necessary because we're using a
'queue' metaphor; we don't guarantee the command has got to the device,
nor do we need to guarantee the command has got to the controller.
Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
During resume, sleep 1 second to wait for the HBA reset
to finish is a waste of time.
According to the AHCI 1.2 spec,
We should poll the HOST_CTL register,
and return error if the host reset is not
finished within 1 second.
Test results show that the HBA reset can be done quickly(in usecs).
And this patch may save nearly 1 second during resume.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
SCSI VPD page b1 reports the nominal rotation speed and physical size
of the device. Devices that conform to ATA-8 can return this information
in words 217 and 168 of the identify data.
Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
drivers/ata/pata_qdi.c:142:9: warning: incorrect type in assignment (different base types)
drivers/ata/pata_qdi.c:142:9: expected unsigned int [unsigned] [usertype] pad
drivers/ata/pata_qdi.c:142:9: got restricted __le32 [usertype] <noident>
drivers/ata/pata_qdi.c:146:15: warning: cast to restricted __le32
drivers/ata/pata_winbond.c:110:9: warning: incorrect type in assignment (different base types)
drivers/ata/pata_winbond.c:110:9: expected unsigned int [unsigned] [usertype] pad
drivers/ata/pata_winbond.c:110:9: got restricted __le32 [usertype] <noident>
drivers/ata/pata_winbond.c:114:15: warning: cast to restricted __le32
drivers/ata/pata_legacy.c:310:9: warning: incorrect type in assignment (different base types)
drivers/ata/pata_legacy.c:310:9: expected unsigned int [unsigned] [usertype] pad
drivers/ata/pata_legacy.c:310:9: got restricted __le32 [usertype] <noident>
drivers/ata/pata_legacy.c:314:15: warning: cast to restricted __le32
drivers/ata/pata_legacy.c:752:11: warning: cast to restricted __le32
drivers/ata/pata_legacy.c:756:9: warning: incorrect type in assignment (different base types)
drivers/ata/pata_legacy.c:756:9: expected unsigned int [unsigned] [addressable] [assigned] [usertype] pad
drivers/ata/pata_legacy.c:756:9: got restricted __le32 [usertype] <noident>
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Update atapi_eh_request_sense() to take @dev, @sense_buf and
@dfl_sense_key instead of taking @qc and extracting information from
it. This change is to make the function more generic and allow it to
be called from other places.
While at it, make cdb initialization use initializer.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Add Enclosure Management support to libata and ahci.
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
ATA_TMOUT_INTERNAL which was 30secs were used for all internal
commands which is way too long when something goes wrong. This patch
implements command type based stepped timeouts. Different command
types can use different timeouts and each command type can use
different timeout values after timeouts.
ie. the initial timeout is set to a value which should cover most of
the cases but not too long so that run away cases don't delay things
too much. After the first try times out, the second try can use
longer timeout and if that one times out too, it can go for full 30sec
timeout.
IDENTIFYs use 5s - 10s - 30s timeout and all other commands use 5s -
10s timeouts.
This patch significantly cuts down the needed time to handle failure
cases while still allowing libata to work with nut job devices through
retries.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
This doesn't introduce any functional changes. This is to make reset
timeout table consistent with to-be-added command timeout tables.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>