This patch deprecates the existing use of an arbitrary value TFRC_SMALLEST_P
for low-threshold values of p. This avoids masking low-resolution errors.
Instead, the code now checks against real boundaries (implemented by preceding
patch) and provides warnings whenever a real value falls below the threshold.
If such messages are observed, it is a better solution to take this as an
indication that the lookup table needs to be re-engineered.
Changelog:
----------
This patch
* makes handling all TFRC resolution errors local to the TFRC library
* removes unnecessary test whether X_calc is 'infinity' due to p==0 -- this
condition is already caught by tfrc_calc_x()
* removes setting ccid3hctx_p = TFRC_SMALLEST_P in ccid3_hc_tx_packet_recv
since this is now done by the TFRC library
* updates BUG_ON test in ccid3_hc_tx_no_feedback_timer to take into account
that p now is either 0 (and then X_calc is irrelevant), or it is > 0; since
the handling of TFRC_SMALLEST_P is now taken care of in the tfrc library
Justification:
--------------
The TFRC code uses a lookup table which has a bounded resolution.
The lowest possible value of the loss event rate `p' which can be
resolved is currently 0.0001. Substituting this lower threshold for
p when p is less than 0.0001 results in a huge, exponentially-growing
error. The error can be computed by the following formula:
(f(0.0001) - f(p))/f(p) * 100 for p < 0.0001
Currently the solution is to use an (arbitrary) value
TFRC_SMALLEST_P = 40 * 1E-6 = 0.00004
and to consider all values below this value as `virtually zero'. Due to
the exponentially growing resolution error, this is not a good idea, since
it hides the fact that the table can not resolve practically occurring cases.
Already at p == TFRC_SMALLEST_P, the error is as high as 58.19%!
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
This
* adds documentation about the lowest resolution that is possible within
the bounds of the current lookup table
* defines a constant TFRC_SMALLEST_P which defines this resolution
* issues a warning if a given value of p is below resolution
* combines two previously adjacent if-blocks of nearly identical
structure into one
This patch does not change the algorithm as such.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
1) For the forward X_calc lookup, it
* protects effectively against RTT=0 (this case is possible), by
returning the maximal lookup value instead of just setting it to 1
* reformulates the array-bounds exceeded condition: this only happens
if p is greater than 1E6 (due to the scaling)
* the case of negative indices can now with certainty be excluded,
since documentation shows that the formulas are within bounds
* additional protection against p = 0 (would give divide-by-zero)
2) For the reverse lookup, it warns against
* protects against exceeding array bounds
* now returns 0 if f(p) = 0, due to function definition
* warns about minimal resolution error and returns the smallest table
value instead of p=0 [this would mask congestion conditions]
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
This fixes the following small error in tfrc_calc_x_reverse_lookup.
1) The table is generated by the following equations:
lookup[index][0] = g((index+1) * 1000000/TFRC_CALC_X_ARRSIZE);
lookup[index][1] = g((index+1) * TFRC_CALC_X_SPLIT/TFRC_CALC_X_ARRSIZE);
where g(q) is 1E6 * f(q/1E6)
2) The reverse lookup assigns an entry in lookup[index][small]
3) This index needs to match the above, i.e.
* if small=0 then
p = (index + 1) * 1000000/TFRC_CALC_X_ARRSIZE
* if small=1 then
p = (index+1) * TFRC_CALC_X_SPLIT/TFRC_CALC_X_ARRSIZE
These are exactly the changes that the patch makes; previously the code did
not conform to the way the lookup table was generated (this difference resulted
in a mean error of about 1.12%).
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
This adds documentation for the TCP Reno throughput equation which is at
the heart of the TFRC sending rate / loss rate calculations.
It spells out precisely how the values were determined and what they mean.
The equations were derived through reverse engineering and found to be
fully accurate (verified using test programs).
This patch does not change any code.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
This avoids a (harmless) warning message being printed at the DCCP server
(the receiver of a DCCP half connection).
Incoming packets are both directed to
* ccid_hc_rx_packet_recv() for the server half
* ccid_hc_tx_packet_recv() for the client half
The message gets printed since on a server the client half is currently not
sending data packets.
This is resolved for the moment by checking the DCCP-role first. In future
times (bidirectional DCCP connections), this test may have to be more
sophisticated.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
The main object of this patch is the following bug:
==> In ccid3_hc_tx_packet_recv, the parameters p and X_recv were updated
_after_ the send rate was calculated. This is clearly an error and is
resolved by re-ordering statements.
In addition,
* r_sample is converted from u32 to long to check whether the time difference
was negative (it would otherwise be converted to a large u32 value)
* protection against RTT=0 (this is possible) is provided in a further patch
* t_elapsed is also converted to long, to match the type of r_sample
* adds a a more debugging information regarding current send rates
* various trivial comment/documentation updates
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Acked-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
This bug resulted in ccid3_hc_tx_send_packet returning negative
delay values, which in turn triggered silently dequeueing packets in
dccp_write_xmit. As a result, only a few out of the submitted packets made
it at all onto the network. Occasionally, when dccp_wait_for_ccid was
involved, this also triggered a bug warning since ccid3_hc_tx_send_packet
returned a negative value (which in reality was a negative delay value).
The cause for this bug lies in the comparison
if (delay >= hctx->ccid3hctx_delta)
return delay / 1000L;
The type of `delay' is `long', that of ccid3hctx_delta is `u32'. When comparing
negative long values against u32 values, the test returned `true' whenever delay
was smaller than 0 (meaning the packet was overdue to send).
The fix is by casting, subtracting, and then testing the difference with
regard to 0.
This has been tested and shown to work.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
The TFRC nofeedback timer normally expires after the maximum of 4
RTTs and twice the current send interval (RFC 3448, 4.3). On LANs
with a small RTT this can mean a high processing load and reduced
performance, since then the nofeedback timer is triggered very
frequently.
This patch provides a configuration option to set the bound for the
nofeedback timer, using as default 100 milliseconds.
By setting the configuration option to 0, strict RFC 3448 behaviour
can be enforced for the nofeedback timer.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
Misc Fixes:
- Prevent references to NULL node list element in reset routines.
- Add missing IOCB types to switch tables
- Reset the card on Port Error 5
- Fix infinite loop in LUN reset
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
The driver now allows both wwpn and wwnn to be set.
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
To avoid continually updating the driver for new subsystem ids
(as adapter modules are proliferating), remove this 2nd level decode.
Genericize the reported Adapter names to be consistent across
Emulex product line.
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Add MSI (Message Signalled Interrupts) support
Actual use must be enabled via the new module parameter "lpfc_use_msi"
Defaults to no use
Many thanks to Frederic Temporelli who implemented the initial patch.
Signed-off-by: Frederic Temporelli <frederic.temporelli@ext.bull.net>
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Adjust LOG_FCP logging to be more meaningful.
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
It was not accounted for in the fast/slow rings.
Genericize the implementation and control it via sysfs
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Discovery Fixes:
- Prevent starting discovery of a node if discovery is in progress.
- Code improvement (reduction) for lpfc_findnode_did().
- Update discovery to send RFF to Fabric on link up
- Bypass unique WWN checks for fabric addresses
- Add ndlp to plogi list prior to issuing the plogi els command
Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
This is IBM Virtual SCSI target driver for tgt. The driver is based on
the original ibmvscsis driver:
http://lkml.org/lkml/2005/10/17/99
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: Santiago Leon <santil@us.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
libsrp provides helper functions for SRP target drivers.
Some SRP target drivers would be out of drivers/scsi/ so we added an
entry for libsrp in drivers/scsi/Kconfig.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: Santiago Leon <santil@us.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
In the switch over, I forgot to set the command length, so it sends out
a request sense with whatever length the prior command had (and fails
badly if it wasn't 6).
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Use NULL instead of 0 for pointers (sparse warning):
drivers/scsi/qla2xxx/qla_attr.c:393:4: warning: Using plain integer as NULL pointer
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
This patch adds code to fix up the PHYMODE4 "align timing"
register value on second-generation Promise SATA chips.
Failure to correct this value on non-x86 machines makes
drive detection prone to failure due to timeouts. (I've
observed about 50% detection failure rates on SPARC64.)
The HW boots with a bad value in this register, but on x86
machines the Promise BIOS corrects it to the value recommended
by the manual, so most people have been unaffected by this issue.
After developing the patch I checked Promise's SATAII driver,
and discovered that it also corrects PHYMODE4 just like this
patch does.
This patch depends on the sata_promise SATAII updates
patch I sent recently.
Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
libata switched to IRQ-driven IDENTIFY when IRQ-driven PIO was
introduced. This has caused a lot of problems including device
misdetection and phantom device.
ATA_FLAG_DETECT_POLLING was added recently to selectively use polling
IDENTIFY on problemetic drivers but many controllers and devices are
affected by this problem and trying to adding ATA_FLAG_DETECT_POLLING
for each such case is diffcult and not very rewarding.
This patch makes libata always use polling IDENTIFY. This is
consistent with libata's original behavior and drivers/ide's behavior.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
ahci_init_controller() calls ahci_deinit_port() to make sure the
controller is stopped before initializing the controller. In turn,
ahci_deinit_port() invokes ahci_power_down() to power down the port.
If the controller supports slumber mode, the link is put into it.
Unfortunately, some devices don't implement link powersaving mode
properly and show erratic behavior after link is put into slumber
mode. For example, HL-DT-ST DVD-RAM GSA-H30N completely locks up on
slumber transition and can only be recovered with the *REAL* hard
reset - power removal and reapply.
Note that this makes the first probing reset different from all
others. If the above dvd-ram is hotplugged after ahci is initialized,
no problem occurs because ahci is already fully initialized with phy
powered up. So, this might also be the reason for other weird AHCI
initial probing abnormalities.
This patch moves power up/down out of port init/deinit and call them
only when needed.
Power down is now called only when suspending. As system suspend
usually involves powering down 12v for storage devices, this shouldn't
cause problem even if the attached device doesn't support slumber
mode. However, in partial power management and suspend failure cases,
devices might lock up after suspend attempt. I thought about removing
transition to slumber mode altogether but ahci spec mandates it before
HBA D3 state transition. Blacklisting such devices might be the
solution.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Separate out rw ATA taskfile building from ata_scsi_rw_xlat() into
ata_build_rw_tf(). This will be used to improve media error handling.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Sg'ify ata_exec_internal() and call it ata_exec_internal_sg().
Wrapper function around ata_exec_internal_sg() is implemented to
provide ata_exec_internal() interface.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Now that BMDMA status is recorded in irq handler. ata_bmdma_freeze()
is free to manipulate host status. Under certain circumstances, some
controllers (ICH7 in enhanced mode w/ IRQ shared) raise IRQ when CTL
register is written to and ATA_NIEN doesn't mask it.
This patch makes ata_bmdma_freeze() clear all pending IRQs after
freezing a port. This change makes explicit clearing in
ata_device_add() unnecessary and thus kills it. The removed code was
SFF-specific and was in the wrong place.
Note that ->freeze() handler is always called under ap->lock held and
irq disabled. Even if CTL manipulation causes stuck IRQ, it's cleared
immediately. This should be safe (enough) even in SMP environment.
More correct solution is to mask the IRQ from IRQ controller but that
would be an overkill.
Signed-off-by: Tejun Heo <htejun@gmail.com>
For certain errors, interrupt handler alter BMDMA host status before
entering EH (clears active and intr). Thus altered BMDMA host status
value is recorded by BMDMA EH and reported to user. Move BMDMA host
status recording from EH to interrupt handler.
Signed-off-by: Tejun Heo <htejun@gmail.com>
ata_scsi_dev_rescan() doesn't synchronize against SCSI device detach
and the target sdev might go away in the middle. Fix it.
Signed-off-by: Tejun Heo <htejun@gmail.com>
If EH command is issued to a frozen port, it fails with AC_ERR_SYSTEM.
libata used to request sense even when the port is frozen needlessly
adding AC_ERR_SYSTEM to err_mask. Don't do it.
Signed-off-by: Tejun Heo <htejun@gmail.com>
* READ CAPACITY (16) implementation fixed. Result was shifted by two
bytes. Carlos Pardo spotted this problem and submitted preliminary
patch. Capacity => 2TB is handled correctly now. (verifid w/ fake
capacity)
* Use dev->n_sectors instead of re-reading directly from ID data.
* Define and use ATA_SCSI_RBUF_SET() which considers rbuf length.
This should be done for all simulation functions. Userland can
issue any simulated command with arbitrary buffer length.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Carlos Pardo <Carlos.Pardo@siliconimage.com>
This patch implements ATA_FLAG_SETXFER_POLLING and use in pata_via.
If this flag is set, transfer mode setting performed by polling not by
interrupt. This should help those controllers which raise interrupt
before the command is actually complete on SETXFER.
Rationale for this approach.
* uses existing facility and relatively simple
* no busy sleep in the interrupt handler
* updating drivers is easy
While at it, kill now unused flag ATA_FLAG_SRST in pata_via.
Signed-off-by: Tejun Heo <htejun@gmail.com>
There are machines out there which share legacy PCI IDE IRQs w/ other
devices. libata SFF interrupt/HSM code is ready for shared IRQ and
has been setting IRQF_SHARED for devices in native PCI mode. Device
in legacy mode is still a PCI device and thus supposedly uses
active-low level triggered IRQ.
Machines with such setup should be quite rare and w/o this flag libata
is likely to fail loading and render the system unuseable. Also, IDE
driver has been setting IRQF_SHARED for devices in legacy mode for a
looooong time.
Signed-off-by: Tejun Heo <htejun@gmail.com>
HSM_ST_UNKNOWN is not used anywhere. Its value is zero and supposed
to serve sanity check purpose but HSM_ST_IDLE is used for that
purpose. This unused state causes confusion. After a port is
initialized but before the first command is executed, the idle hsm
state is UNKNOWN. However, once a command has completed, the idle hsm
state is IDLE. This defeats sanity check in ata_pio_task() for the
first command.
This patch removes HSM_ST_UNKNOWN and consequently make HSM_ST_IDLE
the default state.
Signed-off-by: Tejun Heo <htejun@gmail.com>
sht->max_sectors is overrided unconditionally in ->slave_configure.
There's no reason to set it to any value.
Signed-off-by: Tejun Heo <htejun@gmail.com>
aevents can not uniquely identify an SA. We break the ABI with this
patch, but consensus is that since it is not yet utilized by any
(known) application then it is fine (better do it now than later).
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
To use ipv6_find_hdr(), IP6_NF_IPTABLES is necessary.
Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
David Binderman's icc logs:
net/rose/rose_route.c(399): remark #593: variable "err" was set but never used
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- move EXPORT_SYMBOL next to exported symbol
- use EXPORT_SYMBOL_GPL since this is what the original code used
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Also remove the references to "new connection tracking" from Kconfig.
After some short stabilization period of the new connection tracking
helpers/NAT code the old one will be removed.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add IPv4 and IPv6 capable nf_conntrack port of the TFTP conntrack/NAT helper.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add IPv4 and IPv6 capable nf_conntrack port of the SIP conntrack/NAT helper.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add nf_conntrack port of the PPtP conntrack/NAT helper. Since there seems
to be no IPv6-capable PPtP implementation the helper only support IPv4.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add nf_conntrack port of the NetBIOS name service conntrack helper.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gerrit Renker