Region Flush has a weird programming model.
1. Flush or Invalidate is selected by DC_CTRL.RGN_OP
2 Flush-n-Invalidate is done by DC_CTRL.IM
Given the code structuring before, case #2 above was generating two
seperate updates to DC_CTRL which was pointless.
| 80a342b0 <__dma_cache_wback_inv_l1>:
| 80a342b0: clri r4
| 80a342b4: lr r2,[dc_ctrl]
| 80a342b8: bset_s r2,r2,0x6
| 80a342ba: sr r2,[dc_ctrl] <-- FIRST
|
| 80a342be: bmskn r3,r0,0x5
|
| 80a342c2: lr r2,[dc_ctrl]
| 80a342c6: and r2,r2,0xfffff1ff
| 80a342ce: bset_s r2,r2,0x9
| 80a342d0: sr r2,[dc_ctrl] <-- SECOND
|
| 80a342d4: add_s r1,r1,0x3f
| 80a342d6: bmsk_s r0,r0,0x5
| 80a342d8: add_s r0,r0,r1
| 80a342da: add_s r0,r0,r3
| 80a342dc: sr r0,[78]
| 80a342e0: sr r3,[77]
|...
|...
So move setting of DC_CTRL.RGN_OP into __before_dc_op() and combine with
any other update.
| 80b63324 <__dma_cache_wback_inv_l1>:
| 80b63324: clri r3
| 80b63328: lr r2,[dc_ctrl]
| 80b6332c: and r2,r2,0xfffff1ff
| 80b63334: or r2,r2,576
| 80b63338: sr r2,[dc_ctrl]
|
| 80b6333c: add_s r1,r1,0x3f
| 80b6333e: bmskn r2,r0,0x5
| 80b63342: add_s r0,r0,r1
| 80b63344: sr r0,[78]
| 80b63348: sr r2,[77]
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
As reported in STAR 9001165532, an SLC control reg read (for checking
busy state) right after SLC invalidate command may incorrectly return
NOT busy causing software to NOT spin-wait while operation is underway.
(and for some reason this only happens if L1 cache is also disabled - as
required by IOC programming model)
Suggested workaround is to do an additional Control Reg read, which
ensures the 2nd read gets the right status.
Cc: stable@vger.kernel.org #4.10
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
[vgupta: reworte changelog a bit]
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
The programming model has been fixed with prev patches so re-enable it
by default
This reverts commit 23cb1f6440.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
arc_cache_init() is called for each core so can't be tagged __init.
However bulk of it is only executed by master core and thus is candidate
for __init reaping.
So split it up to allow that.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
vs. fixed 512M before.
But this still assumes that all of memory is under IOC which may not be
true for the SoC. Improve that later when this becomes a real issue, by
specifying this from DT.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
On AXS103 release bitfiles, DMA data corruptions were seen because IOC
setup was not following the recommended way in documentation.
Flipping IOC on when caches are enabled or coherency transactions are in
flight, might cause some of the memory operations to not observe
coherency as expected.
So strictly follow the programming model recommendations as documented
in comment header above arc_ioc_setup()
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
An ARC700 customer reported linux boot crashes when upgrading to bigger
L1 dcache (64K from 32K). Turns out they had an aliasing VIPT config and
current code only assumed 2 colours, while theirs had 4. So default to 4
colours and complain if there are fewer. Ideally this needs to be a
Kconfig option, but heck that's too much of hassle for a single user.
Cc: stable@vger.kernel.org
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Historical MMU revisions have been paired with Cache revision updates
which are captured in MMU and Cache Build Configuration Registers respectively.
This was used in boot code to check for configurations mismatches,
speically in simulations (such as running with non existent caches,
non pairing MMU and Cache version etc). This can instead be inferred
from other cache params such as line size. So remove @ver from post
processed @cpuinfo which could be used later to save soem other
interesting info.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Previously we would not print the case when IOC existed but was not
enabled.
And while at it, reduce one line off boot printing by consolidating
the Peripheral address space and IO-Coherency which in a way
applies to them
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
if user disables IOC from debugger at startup (by clearing @ioc_enable),
@ioc_exists is cleared too. This means boot prints don't capture the
fact that IOC was present but disabled which could be misleading.
So invert how we use @ioc_enable and @ioc_exists and make it more
canonical. @ioc_exists represent whether hardware is present or not and
stays same whether enabled or not. @ioc_enable is still user driven,
but will be auto-disabled if IOC hardware is not present, i.e. if
@ioc_exist=0. This is opposite to what we were doing before, but much
clearer.
This means @ioc_enable is now the "exported" toggle in rest of code such
as dma mapping API.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
HS release 3.0 provides for even more flexibility in specifying the
volatile address space for mapping peripherals.
With HS 2.1 @start was made flexible / programmable - with HS 3.0 even
@end can be setup (vs. fixed to 0xFFFF_FFFF before).
So add code to reflect that and while at it remove an unused struct
defintion
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
For resources shared by all cores such as SLC and IOC, only the master
core needs to do any setups / enabling / disabling etc.
Cc: <stable@vger.kernel.org>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The peripheral address space is architectural address window which is
uncached and typically used to wire up peripherals.
For ARC700 cores (ARCompact ISA based) this was fixed to 1GB region
0xC000_0000 - 0xFFFF_FFFF.
For ARCv2 based HS38 cores the start address is flexible and can be
0xC, 0xD, 0xE, 0xF 000_000 by programming AUX_NON_VOLATILE_LIMIT reg
(typically done in bootloader)
Further in cas of PAE, the physical address can extend beyond 4GB so
need to confine this check, otherwise all pages beyond 4GB will be
treated as uncached
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Let's define page_mapped() to be true for compound pages if any
sub-pages of the compound page is mapped (with PMD or PTE).
On other hand page_mapcount() return mapcount for this particular small
page.
This will make cases like page_get_anon_vma() behave correctly once we
allow huge pages to be mapped with PTE.
Most users outside core-mm should use page_mapcount() instead of
page_mapped().
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Tested-by: Sasha Levin <sasha.levin@oracle.com>
Tested-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Acked-by: Jerome Marchand <jmarchan@redhat.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Steve Capper <steve.capper@linaro.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is the first working implementation of 40-bit physical address
extension on ARCv2.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
That way a single flip of phys_addr_t to 64 bit ensures all places
dealing with physical addresses get correct data
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Before we plug in highmem support, some of code needs to be ready for it
- copy_user_highpage() needs to be using the kmap_atomic API
- mk_pte() can't assume page_address()
- do_page_fault() can't assume VMALLOC_END is end of kernel vaddr space
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
In case of ARCv2 CPU there're could be following configurations
that affect cache handling for data exchanged with peripherals
via DMA:
[1] Only L1 cache exists
[2] Both L1 and L2 exist, but no IO coherency unit
[3] L1, L2 caches and IO coherency unit exist
Current implementation takes care of [1] and [2].
Moreover support of [2] is implemented with run-time check
for SLC existence which is not super optimal.
This patch introduces support of [3] and rework of DMA ops
usage. Instead of doing run-time check every time a particular
DMA op is executed we'll have 3 different implementations of
DMA ops and select appropriate one during init.
As for IOC support for it we need:
[a] Implement empty DMA ops because IOC takes care of cache
coherency with DMAed data
[b] Route dma_alloc_coherent() via dma_alloc_noncoherent()
This is required to make IOC work in first place and also
serves as optimization as LD/ST to coherent buffers can be
srviced from caches w/o going all the way to memory
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
[vgupta:
-Added some comments about IOC gains
-Marked dma ops as static,
-Massaged changelog a bit]
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
SLC maintenance ops need to be serialized by software as there is no
inherent buffering / quequing of aux commands. It can silently ignore a
new aux operation if previous one is still ongoing (SLC_CTRL_BUSY)
So gaurd the SLC op using a spin lock
The spin lock doesn't seem to be contended even in heavy workloads such
as iperf. On FPGA @ 75 MHz.
[1] Before this change:
============================================================
# iperf -c 10.42.0.1
------------------------------------------------------------
Client connecting to 10.42.0.1, TCP port 5001
TCP window size: 43.8 KByte (default)
------------------------------------------------------------
[ 3] local 10.42.0.110 port 38935 connected with 10.42.0.1 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 48.4 MBytes 40.6 Mbits/sec
============================================================
[2] After this change:
============================================================
# iperf -c 10.42.0.1
------------------------------------------------------------
Client connecting to 10.42.0.1, TCP port 5001
TCP window size: 43.8 KByte (default)
------------------------------------------------------------
[ 3] local 10.42.0.243 port 60248 connected with 10.42.0.1 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 47.5 MBytes 39.8 Mbits/sec
# iperf -c 10.42.0.1
------------------------------------------------------------
Client connecting to 10.42.0.1, TCP port 5001
TCP window size: 43.8 KByte (default)
------------------------------------------------------------
[ 3] local 10.42.0.243 port 60249 connected with 10.42.0.1 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 54.9 MBytes 46.0 Mbits/sec
============================================================
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: arc-linux-dev@synopsys.com
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
L2 cache on ARCHS processors is called SLC (System Level Cache)
For working DMA (in absence of hardware assisted IO Coherency) we need
to manage SLC explicitly when buffers transition between cpu and
controllers.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Caveats about cache flush on ARCv2 based cores
- dcache is PIPT so paddr is sufficient for cache maintenance ops (no
need to setup PTAG reg
- icache is still VIPT but only aliasing configs need PTAG setup
So basically this is departure from MMU-v3 which always need vaddr in
line ops registers (DC_IVDL, DC_FLDL, IC_IVIL) but paddr in DC_PTAG,
IC_PTAG respectively.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
- Remove the ifdef'ery and write distinct versions for each mmu ver even
if there is some code duplication
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
That is because __after_dc_op() already reads it for status check, so it
is better anyways to use that "newer" value.
Also reduces the clutter in callers for passing from/to these routines.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
