Commit Graph

1243 Commits

Author SHA1 Message Date
Adrian Chadd
3f08db2e79 Protect the PCI space registers behind a mutex.
Obtained from:	Linux/OpenWRT, Atheros
2012-04-17 01:22:59 +00:00
Adrian Chadd
468d6f48b3 Add in the AP96 phy configuration from openwrt.
* arge0 doesn't (yet) work via the switch PHY ports; I'm not sure why.
* arge1 maps to the WAN port. That works.

TODO:

* The PLL register needs a different (non-default) value for Gigabit
  Ethernet.  The board setup code needs to be extended a bit to allow
  for non-default pll_1000 values - right now, those values come out
  of hard-coded values in the per-chip set_pll_ge() routines.

Obtained from:	Linux / OpenWRT
2012-04-15 22:59:56 +00:00
Adrian Chadd
5fdb2379cb The AR913x MII speed configuration matches the AR71xx MII configuration.
So share the code.

Don't do it for the AR724x - that has a completely different set of PLL
and MII configuration parameters.
2012-04-15 22:34:22 +00:00
Adrian Chadd
bf9abaa954 Fix the mask logic when reading PCI configuration space registers. 2012-04-15 02:38:01 +00:00
Adrian Chadd
c4b28bdc27 Flesh out the rest of the AP96 board/config. 2012-04-13 20:23:32 +00:00
Adrian Chadd
d591b27dbc * Enable ATH_EEPROM_FIRMWARE, now that it's a compile time option
* Tidy up things a bit.
2012-04-13 18:01:53 +00:00
Adrian Chadd
2c61ba4db2 These are uboot, so mark them as such or booting from flash will not work. 2012-04-13 08:56:23 +00:00
Adrian Chadd
3a8a3eebfd Introduce configuration files for AP94 and AP96.
This uses the new firmware(9) method for squirreling away the EEPROM
contents from SPI flash so ath(4) can get to them later.

It won't work out of the box just yet - you have to add this to
if_ath_pci.c:

#define ATH_EEPROM_FIRMWARE

.. until I've added it as a configuration option and updated things.
2012-04-13 08:52:25 +00:00
Adrian Chadd
8f7015e205 (ab)Use the firmware API to store away EEPROM calibration data for
future use by the ath(4) driver.

These embedded devices put the calibration/PCI bootstrap data on the
on board SPI flash rather than on an EEPROM connected to the NIC.
For some boards, there's two NICs and two sets of EEPROM data in the
main SPI flash.

The particulars:

* Introduce ath_fixup_size, which is the size of the EEPROM area in
  bytes.
* Create a firmware image with a name based on the PCI device identifier
  (bus/slot/device/function).
* Hide some verbose debugging behind 'bootverbose'.

ath(4) can then use this to load in the EEPROM data.

This requires AR71XX_ATH_EEPROM to be defined.
2012-04-13 08:45:50 +00:00
Adrian Chadd
8a138d80d0 Remove an unused variable. Grr. 2012-04-13 06:13:37 +00:00
Adrian Chadd
be94a28e2a Sync this code against what's in OpenWRT trunk.
* the openwrt code doesn't treat 0/0/0 any differently
  from other bus/slot/func combinations.
* A "local write" function writes to the LCONF area, and
  so I've added it.
* The PCI workaround at attach time uses this LCONF code,
  which it already did ..
* .. but it is a 4 byte write, not a 2 byte write.
  Even though it's PCIR_COMMAND which is a two byte PCI register.

Tested on:	AR7161
TODO:		The other two AR71xx derivatives
TODO:		More thoroughly stare at the datasheets I do have
		and if it indeed is incorrect, push fixes to both
		FreeBSD and Linux/OpenWRT.

Obtained from:	Linux OpenWRT
2012-04-13 06:11:24 +00:00
Jayachandran C.
2323b34a2d Reinstate the XTLB handler for CPU_NLM and CPU_RMI
These platforms set the KX bit even when booted in 32 bit mode. So
the XLTB handler is needed even when __mips_n64 is not defined.
2012-04-02 11:41:33 +00:00
John Baldwin
5e1a7cc71e Use VM_MEMATTR_UNCACHEABLE for the constant for UC memory rather than
VM_MEMATTR_UNCACHED.  VM_MEMATTR_UNCACHEABLE is the constant other
platforms use.

MFC after:	2 weeks
2012-03-29 16:48:36 +00:00
Jayachandran C.
7fb26c47df Remove unnecessary assembly code.
The compiler should generate lw/sw corresponding to register
operations.
2012-03-29 11:46:29 +00:00
Juli Mallett
84db023ec1 Assume a big-endian default on MIPS and drop the "eb" suffix from MACHINE_ARCH.
This makes our naming scheme more closely match other systems and the
expectations of much third-party software.  MIPS builds which are little-endian
should require and exhibit no changes.  Big-endian TARGET_ARCHes must be
changed:
	From:		To:
	mipseb		mips
	mipsn32eb	mipsn32
	mips64eb	mips64

An entry has been added to UPDATING and some foot-shooting protection (complete
with warnings which should become errors in the near future) to the top-level
base system Makefile.
2012-03-29 02:54:35 +00:00
Juli Mallett
df42d19401 Turn on messages from the Simple Executive codebase, what few there are. 2012-03-29 02:05:11 +00:00
Juli Mallett
5143d82211 Disable FP instruction emulation by default on !o32 because of ABI concerns.
Note that in practice this isn't needed because we get a coprocessor unusable
exception first, but that's actually something like a bug.
2012-03-29 02:04:15 +00:00
Juli Mallett
39dec33f2b Supply endianness implied by the -m flag when compiling ucore code. 2012-03-29 02:03:06 +00:00
Juli Mallett
5c3c01764b Fix little-endian built. 2012-03-29 02:02:23 +00:00
Fabien Thomas
f5f9340b98 Add software PMC support.
New kernel events can be added at various location for sampling or counting.
This will for example allow easy system profiling whatever the processor is
with known tools like pmcstat(8).

Simultaneous usage of software PMC and hardware PMC is possible, for example
looking at the lock acquire failure, page fault while sampling on
instructions.

Sponsored by: NETASQ
MFC after:	1 month
2012-03-28 20:58:30 +00:00
Jayachandran C.
eff39d0f38 Fix size of PCI softc. 2012-03-27 18:26:35 +00:00
Jayachandran C.
4b3aada9d4 Resource allocation for XLP SoC SDHCI slots
The on-chip SD slots do not have PCI BARs corresponding to them, so
this has to be handled in the custom SoC memory allocation.

Provide memory resource for rids corresponding to BAR 0 and 1 in
the custom allocation code.
2012-03-27 15:43:32 +00:00
Jayachandran C.
d5d4261f35 Update memory and resource allocation code for SoC devices
The XLP on-chip devices have PCI configuration headers, but some of the
devices need custom resource allocation code.
- devices with no MEM/IO BARs with registers in PCIe extended reg
  space have to be handled in memory resource allocation
- devices without INTPIN/INTLINE in PCI header can be supported
  by having these faked with a shadow register.
- Some devices does not allow 8/16 bit access to the register space,
  he default bus space cannot be used for these.

Subclass pci and override attach and resource allocation methods to
take care of this.

Remove earlier code which did this partially.
2012-03-27 15:39:55 +00:00
Jayachandran C.
2652f84c92 NOR flash driver for XLP.
The NOR interface on the SoC appears on the top level PCI bus. Add
a simple driver for this.
2012-03-27 15:16:38 +00:00
Jayachandran C.
250a191260 XLP UART code udpate.
Move XLP PCI UART device to sys/mips/nlm/dev/ directory.  Other
drivers for the XLP SoC devices will be added here as well.
Update uart_cpu_xlp.c and uart_pci_xlp.c use macros for uart port,
speed and IO frequency.
2012-03-27 14:48:40 +00:00
Jayachandran C.
35011d20cb xlpge : driver for XLP network accelerator
Features:
- network driver for the four 10G interfaces and two management ports
  on XLP 8xx.
- Support 4xx and 3xx variants of the processor.
- Source code and firmware building for the 16 mips32r2 micro-code engines
  in the Network Accelerator.
- Basic initialization code for Packet ordering Engine.

Submitted by:	Prabhath Raman (prabhath at netlogicmicro com)
		[refactored and fixed up for style by jchandra]
2012-03-27 14:05:12 +00:00
Jayachandran C.
ca950537bb Support for EEPROM and CPLD on XLP EVP boards.
On XLP evaluation platform, the board information is stored
in an I2C eeprom and the network block configuration is available
from a CPLD connected to the GBU (NOR flash bus). Add support
for both of these.
2012-03-27 12:25:47 +00:00
Jayachandran C.
9b4d140639 Opencrypto driver for XLP Security and RSA/ECC blocks
Support for the Security and RSA blocks on XLP SoC. Even though
the XLP supports many more algorithms, only the ones supported
in OCF have been added.

Submitted by:	Venkatesh J. V. (venkatesh at netlogicmicro com)
2012-03-27 11:43:46 +00:00
Jayachandran C.
8f57f9e0d7 I2C support for XLP, add hints for I2C devices and update PCI resource
allocation code.
2012-03-27 11:17:04 +00:00
Jayachandran C.
e12ed67f35 XLP PCIe code update.
- XLP supports hardware swap for PCIe IO/MEM accesses. Since we
  are in big-endian mode, enable hardware swap and use the normal
  bus space.
- move some printfs to bootverbose, and remove others.
- fix SoC device resource allocation code
- Do not use '|' while updating PCIE_BRIDGE_MSI_ADDRL
- some style fixes

In collaboration with: Venkatesh J. V. (venkatesh at netlogicmicro com)
2012-03-27 07:57:41 +00:00
Jayachandran C.
752a2fe776 Update the L1D cache flush sequence when enabling threads.
Added more comments to the code.
2012-03-27 07:51:42 +00:00
Jayachandran C.
68790561a5 Switch to interrupt based message handling for XLP 8xx B0.
Fixup some style issues in the file as well.
2012-03-27 07:47:13 +00:00
Jayachandran C.
91339fd498 Support for XLP4xx and XLP 8xx B0 revision
- Add 4xx processor IDs, add workaround in CPU detection code.
- Update frequency detection code for XLP 8xx.
- Add setting device frequency code.
- Update processor ID checking code.
2012-03-27 07:39:05 +00:00
Jayachandran C.
7aeda475e5 Fixes to the XLP startup code.
Changes are:
- Correct the order of calling init functions.
- Fix up checking excluding reset area.
2012-03-27 07:34:27 +00:00
Oleksandr Tymoshenko
5e608d0b09 Remap PMC interrupt for all cores 2012-03-24 06:28:15 +00:00
Oleksandr Tymoshenko
fa1e9b3a72 Add DTrace-related part to machine-dependent code:
- DTrace trap handler
- invop-related variables (unused on MIPS but still referenced from dtrace)
2012-03-24 05:17:38 +00:00
Oleksandr Tymoshenko
2a4f7a57fe Fix pmap_kextract prototype to align it with pmap.c change 2012-03-23 18:07:12 +00:00
Oleksandr Tymoshenko
f86a93b5be Add pseudo-device for handling PMC interrupts and link everything
PMC-related to build
2012-03-23 00:11:54 +00:00
Oleksandr Tymoshenko
ae849c7c42 Setup fake MODINFO variables for octeon kernel 2012-03-23 00:01:09 +00:00
Oleksandr Tymoshenko
1b930fb786 Rework MIPS PMC code:
- Replace MIPS24K-specific code with more generic framework that will
    make adding new CPU support easier
- Add MIPS24K support for new framework
- Limit backtrace depth to 1 for stability reasons and add option
    HWPMC_MIPS_BACKTRACE to override this limitation
2012-03-22 18:01:23 +00:00
Oleksandr Tymoshenko
a5fbfee1d0 Move PMC hook invocation to cpu_intr. The idea is the same as with ast()
call but there is no reason to implement it in assembler.
2012-03-22 17:47:52 +00:00
Jayachandran C.
0e37e252d2 Fix return type of mips pmap_kextract()
The return type should be vm_paddr_t, not vm_offset_t.
2012-03-22 15:14:10 +00:00
Juli Mallett
ce30db0b31 o) Don't keep drive parameters in a global, put them in the softc.
o) Don't report a bogus stripesize.
o) Eliminate a nearby gratuitous "proxy" indirection.
2012-03-21 10:27:12 +00:00
Ed Schouten
92396a3174 Remove pty(4) from our kernel configurations.
As of FreeBSD 8, this driver should not be used. Applications that use
posix_openpt(2) and openpty(3) use the pts(4) that is built into the
kernel unconditionally. If it turns out high profile depend on the
pty(4) module anyway, I'd rather get those fixed. So please report any
issues to me.

The pty(4) module is still available as a kernel module of course, so a
simple `kldload pty' can be used to run old-style pseudo-terminals.
2012-03-21 08:38:42 +00:00
Oleksandr Tymoshenko
fce0c7295d - Fix logic for detection if further processing of PMC should be performed.
pmc_intr returns one if one of the counters actually triggered the IRQ
- style(9) fixed
2012-03-18 01:43:41 +00:00
Adrian Chadd
0c7a79d76e style(9) changes. 2012-03-17 07:29:11 +00:00
Adrian Chadd
0e69f431d4 Begin fleshing out MII clock rate configuration changes.
These are needed for some particular port configurations where the default
speed isn't suitable for all link speed types. (Ie, changing 10/100/1000MBit
PLL rate requires a similar MII clock rate, rather than a fixed MII rate.)

This is:

* only currently implemented for the ar71xx;
* isn't used anywhere (yet), as the final interface for this hasn't yet
  been determined.
2012-03-17 07:25:23 +00:00
Oleksandr Tymoshenko
66134c7e59 Clean-up fake preload data generator:
- Use macros to push scalar values
- Fix type mismatch for module size
2012-03-15 23:53:24 +00:00
Oleksandr Tymoshenko
fba09d4c08 Fill out fake preload structure to let userland tools like pmc(3) know
about kernel module base address and actual size
2012-03-15 05:29:51 +00:00
Adrian Chadd
49def78fbb Remove a now unneeded ARGE_UNLOCK().
Whilst I'm here, remove a couple blank lines.
2012-03-13 06:50:56 +00:00