The Tilera hypervisor shipped in releases up through MDE 4.1 launches
the client operating system (i.e. Linux) at privilege level 1 (PL1).
Starting with MDE 4.2, as part of the work to enable KVM, the
Tilera hypervisor launches Linux at PL2 instead.
This commit makes the KERNEL_PL option default to 2 for tilegx, while
still saying at 1 for tilepro, which doesn't have an updated hypervisor.
It also explains how and when you might want to choose another value.
In addition, we change a small buglet in the on-chip Ethernet driver,
where we were failing to use the KERNEL_PL constant in an API call.
To make the transition cleaner, this change also provides the updated
hv_init() API for the new hypervisor that supports announcing Linux's
compiled-in PL, so the hypervisor can generate a suitable error in the
case of a mismatched hypervisor and Linux binary.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Cc: stable@vger.linux.org
Convert #include "..." to #include <path/...> in kernel system headers.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
This change adds support for accessing the USB shim from within the
kernel. Note that this change by itself does not allow the kernel
to act as a host or as a device; it merely exposes the built-in on-chip
hardware to the kernel.
The <arch/usb_host.h> and <arch/usb_host_def.h> headers are empty at
the moment because the kernel does not require any types or definitions
specific to the tilegx USB shim; the generic USB core code is all we need.
The headers are left in as stubs so that we don't need to modify the
hypervisor header (drv_usb_host_intf.h) from upstream.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Provide kernel support for the tilegx "Transaction I/O" (TRIO) on-chip
hardware. This hardware implements the PCIe interface for tilegx;
the driver changes to use TRIO for PCIe are in a subsequent commit.
The change is layered on top of the tilegx GXIO IORPC subsystem.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The TILE-Gx chip includes a packet-processing network engine called
mPIPE ("Multicore Programmable Intelligent Packet Engine"). This
change adds support for using the mPIPE engine from within the
kernel. The engine has more functionality than is exposed here,
but to keep the kernel code and binary simpler, this is a subset
of the full API designed to enable standard Linux networking only.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The GXIO I/O RPC subsystem handles exporting I/O hardware resources to
Linux and to applications running under Linux.
For instance, memory which is made available for I/O DMA must be mapped
by an I/O TLB; that means that such memory must be locked down by Linux,
so that it is not swapped or otherwise reused, as long as those I/O
TLB entries are active. Similarly, configuring direct hardware access
introduces new validation requirements. If a user application registers
memory, Linux must ensure that the supplied virtual addresses are valid,
and turn them into client physical addresses. Similarly, when Linux then
supplies those client physical addresses to the Tilera hypervisor, it
must in turn validate those before turning them into the real physical
addresses which are required by the hardware.
To the extent that these sorts of activities were required on previous
TILE architecture processors, they were implemented in a device-specific
fashion. This meant that every I/O device had its own Tilera hypervisor
driver, its own Linux driver, and in some cases its own user-level
library support. There was a large amount of more-or-less functionally
identical code in different places, particularly in the different Linux
drivers. For TILE-Gx, this support has been generalized into a common
framework, known as the I/O RPC framework or just IORPC.
The two "gxio" directories (one for headers, one for sources) start
with just a few files in each with this infrastructure commit, but
after adding support for the on-board I/O shims for networking, PCI,
USB, crypto, compression, I2CS, etc., there end up being about 20 files
in each directory.
More information on the IORPC framework is in the <hv/iorpc.h> header,
included in this commit.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This just adds a few more attributes to the information Linux
can query from the hypervisor for the /sys/hypervisor/board/ directory,
providing part, serial#, revision#, and description for cpu modules
(as opposed to the board itself, or any mezzanine boards).
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This change adds support for a new "super" bit in the PTE, using the new
arch_make_huge_pte() method. The Tilera hypervisor sees the bit set at a
given level of the page table and gangs together 4, 16, or 64 consecutive
pages from that level of the hierarchy to create a larger TLB entry.
One extra "super" page size can be specified at each of the three levels
of the page table hierarchy on tilegx, using the "hugepagesz" argument
on the boot command line. A new hypervisor API is added to allow Linux
to tell the hypervisor how many PTEs to gang together at each level of
the page table.
To allow pre-allocating huge pages larger than the buddy allocator can
handle, this change modifies the Tilera bootmem support to put all of
memory on tilegx platforms into bootmem.
As part of this change I eliminate the vestigial CONFIG_HIGHPTE support,
which never worked anyway, and eliminate the hv_page_size() API in favor
of the standard vma_kernel_pagesize() API.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This change introduces new flags for the hv_install_context()
API that passes a page table pointer to the hypervisor. Clients
can explicitly request 4K, 16K, or 64K small pages when they
install a new context. In practice, the page size is fixed at
kernel compile time and the same size is always requested every
time a new page table is installed.
The <hv/hypervisor.h> header changes so that it provides more abstract
macros for managing "page" things like PFNs and page tables. For
example there is now a HV_DEFAULT_PAGE_SIZE_SMALL instead of the old
HV_PAGE_SIZE_SMALL. The various PFN routines have been eliminated and
only PA- or PTFN-based ones remain (since PTFNs are always expressed
in fixed 2KB "page" size). The page-table management macros are
renamed with a leading underscore and take page-size arguments with
the presumption that clients will use those macros in some single
place to provide the "real" macros they will use themselves.
I happened to notice the old hv_set_caching() API was totally broken
(it assumed 4KB pages) so I changed it so it would nominally work
correctly with other page sizes.
Tag modules with the page size so you can't load a module built with
a conflicting page size. (And add a test for SMP while we're at it.)
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The toolchain supports big-endian mode now, so add support for building
the kernel to run big-endian as well.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The change adds some infrastructure for managing tile pmd's more generally,
using pte_pmd() and pmd_pte() methods to translate pmd values to and
from ptes, since on TILEPro a pmd is really just a nested structure
holding a pgd (aka pte). Several existing pmd methods are moved into
this framework, and a whole raft of additional pmd accessors are defined
that are used by the transparent hugepage framework.
The tile PTE now has a "client2" bit. The bit is used to indicate a
transparent huge page is in the process of being split into subpages.
This change also fixes a generic bug where the return value of the
generic pmdp_splitting_flush() was incorrect.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The first version of this patch proposed an arch/tile/drivers/ directory,
but the consensus was that this was probably a poor choice for a place to
group Tilera-specific drivers, and that in any case grouping by platform
was discouraged, and grouping by function was preferred.
This version of the patch addresses various issues raised in the
community, primarily the absence of sysfs integration. The sysfs
integration now handles passing information on sector size, page size,
and total partition size to userspace as well. In addition, we now
use a single "struct cdev" to manage all the partition minor devices,
and dynamically discover the correct number of partitions from the
hypervisor rather than using a module_param with a default value.
This driver has no particular "peer" drivers it can be grouped with.
It is sort of like an MTD driver for SPI ROM, but it doesn't group well
with the other MTD devices since it relies on hypervisor virtualization
to handle many of the irritating aspects of flash ROM management: sector
awareness, background read for sub-sector writes, bit examination to
determine whether a sector erase needs to be issued, etc. It is in fact
more like an EEPROM driver, but the hypervisor virtualization does require
a "flush" command if you wish to commit a sector write prior to writing
to a different sector, and this is sufficiently different from generic
I2C/SPI EEPROMs that as a result it doesn't group well with them either.
The simple character device is already in use by a range of Tilera
SPI ROM management tools, as well as by customers. In addition, using
the simple character device actually simplifies the userspace tools,
since they don't need to manage sector erase, background read, etc.
This both simplifies the code (since we can uniformly manage plain files
and the SPI ROM) as well as makes the user code portable to non-Linux
platforms that don't offer the same MTD ioctls.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
User space code has been able to discover the static page size
by including a special <hv/pagesize.h> file. In the current release,
that file is now gone, and <asm/page.h> doesn't rely on it. The
getpagesize() API is now the only way for userspace to get the page size.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Add tile support for the EDAC driver, which provides unified system
error (memory, PCI, etc.) reporting. For now, the TILEPro port
reports memory correctable error (CE) only.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
The current implementations of __ndelay and __udelay call a hypervisor
service to delay, but the hypervisor service isn't actually implemented
very well, and the consensus is that Linux should handle figuring this
out natively and not use a hypervisor service.
By converting nanoseconds to cycles, and then spinning until the
cycle counter reaches the desired cycle, we get several benefits:
first, we are sensitive to the actual clock speed; second, we use
less power by issuing a slow SPR read once every six cycles while
we delay; and third, we properly handle the case of an interrupt by
exiting at the target time rather than after some number of cycles.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This change adds the first network driver for the tile architecture,
supporting the on-chip XGBE and GBE shims.
The infrastructure is present for the TILE-Gx networking drivers (another
three source files in the new directory) but for now the the actual
tilegx sources are waiting on releasing hardware to initial customers.
Note that arch/tile/include/hv/* are "upstream" headers from the
Tilera hypervisor and will probably benefit less from LKML review.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
While not a port to KVM (yet), this change modifies the kernel
to be able to build either at PL1 or at PL2 with a suitable
config switch. Pushing up this change avoids handling branch
merge issues going forward with the KVM work.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This change rolls up random cleanups not representing any actual bugs.
- Remove a stale CONFIG_ value from the default tile_defconfig
- Remove unused tns_atomic_xxx() family of methods from <asm/atomic.h>
- Optimize get_order() using Tile's "clz" instruction
- Fix a bad hypervisor upcall name (not currently used in Linux anyway)
- Use __copy_in_user_inatomic() name for consistency, and export it
- Export some additional hypervisor driver I/O upcalls and some homecache calls
- Remove the obfuscating MEMCPY_TEST_WH64 support code
- Other stray comment cleanups, #if 0 removal, etc.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
None of these changes fix any actual bugs, but are just various cleanups
that fell out along the way. In particular, some unused #defines and
includes are removed, PREFETCH_STRIDE is added (the default is right for
our shipping chips, but wrong for our next generation), our tile-specific
prefetching code is removed so the (identical) generic prefetching code
can be used instead, a comment is fixed to be proper GPL and not just a
"paste GPL here" token, a "//" comment is converted to "/* */", etc.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
This model is based on the on-chip interrupt model used by the
TILE-Gx next-generation hardware, and interacts much more cleanly
with the Linux generic IRQ layer.
The change includes modifications to the Tilera hypervisor, which
are reflected in the hypervisor headers in arch/tile/include/arch/.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
This change is the core kernel support for TILEPro and TILE64 chips.
No driver support (except the console driver) is included yet.
This includes the relevant Linux headers in asm/; the low-level
low-level "Tile architecture" headers in arch/, which are
shared with the hypervisor, etc., and are build-system agnostic;
and the relevant hypervisor headers in hv/.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Reviewed-by: Paul Mundt <lethal@linux-sh.org>
Chris Metcalf
David Howells
Lucas De Marchi