when msdc_cmd_is_ready return fail, the req_timeout work has not been
inited and cancel_delayed_work() will return false, then, the request
return directly and never call mmc_request_done().
so need call mod_delayed_work() before msdc_cmd_is_ready()
Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
as the mmc core layer has the mmc->actual_clock, so fill it
and drop msdc_host->sclk.
Signed-off-by: Chaotian Jing <chaotian.jing@mediatek.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
host->chan_rx is NULL when UNIPHIER_SD_CAP_BROKEN_DMA_RX quirk flag
is set. In this case, it should not set up DMA.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Once DMA is enabled, it is not possible to disable it because
uniphier_sd_dma_endisable() always sets the DMA_ENABLE_DMASDRW bit
regardless of the argument 'enable'. It should disable DMA when
'enable' is false.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
host->chan_{rx,tx} represents the DMA capability of the platform.
Even if DMA is supported, there are cases where we want to use PIO,
for example, data length is short enough as commit 5f52c35529
("mmc: tmio: use PIO for short transfers") mentioned.
Regarding the hardware control flow, we are interested in whether DMA
is currently enabled or not, instead of whether the platform has the
DMA capability.
Hence, the several conditionals in tmio_mmc_core.c end up with
checking host->chan_{rx,tx} and !host->force_pio. This is not nice.
Let's flip the flag host->force_pio into host->dma_on.
host->dma_on represents whether the DMA is currently enabled or not.
This flag is set false in the beginning of each command, then should
be set true by tmio_mmc_start_dma() when the DMA is turned on.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
TMIO_MMC_HAVE_HIGH_REG is confusing due to its counter-intuitive name.
All the TMIO MMC variants (TMIO MMC, Renesas SDHI, UniPhier SD) actually
have high registers. It is just that each of them implements its own
registers there. The original IP from Panasonic only defines registers
0x00-0xff in the bus_shift=1 review. The register area above them is
platform-dependent.
In fact, TMIO_MMC_HAVE_HIGH_REG is set only by tmio-mmc.c and used to
test the accessibility of CTL_SDIO_REGS. Because it is specific to
the TMIO MFD variant, the right thing to do is to move such registers
to tmio_mmc.c and delete the TMIO_MMC_HAVE_HIGH_REG flag.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
CTL_RESET_SDIO register is specific to the TMIO MFD (tmio_mmc.c).
Add a new hook host->reset() for performing a platform-specific
reset sequence, and move CTL_RESET_SDIO over there.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The RZ/G1C (a.k.a. R8A77470) comes with three SDHI interfaces,
SDHI0 and SDHI2 are compatible with the R-Car Gen2 SDHIs, SDHI1
is compatible with R-Car Gen3 SDHIs and it can be used as
eMMC as well. This patch adds driver compatibility, and makes
sure both drivers get compiled for the R8A77470.
Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Biju Das <biju.das@bp.renesas.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The RZ/G1C (a.k.a. R8A77470) comes with three SDHI interfaces,
SDHI0 and SDHI2 are compatible with R-Car Gen2 SDHIs, and
SDHI1 is compatible with R-Car Gen3 SDHIs, as it comes with an
internal DMAC, therefore SDHI1 is fully compatible with driver
renesas_sdhi_internal_dmac driver. As a result, the compatible
strings for the R8A77470 SDHI interfaces are a little bit special.
Document SDHI support for the RZ/G1C SoC.
Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Biju Das <biju.das@bp.renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Since commit 9ec36cafe4 ("of/irq: do irq resolution in platform_get_irq")
platform_get_irq() can return -EPROBE_DEFER. However, the driver overrides
an error returned by that function with -ENOENT which breaks the deferred
probing. Propagate upstream an error code returned by platform_get_irq()
and remove the bogus "platform" from the error message, while at it...
Fixes: 9ec36cafe4 ("of/irq: do irq resolution in platform_get_irq")
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This patch adds da7219_max98927 machine driver entry into
machine table
Signed-off-by: Mac Chiang <mac.chiang@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This patch adds Kabylake I2S machine driver with:
DA7219 audio codec(SSP1) and MAXIM98927(SSP0) speaker amplifier.
Signed-off-by: Mac Chiang <mac.chiang@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Some 3 pole connectors report impedance greater than threshold of
1000Ohm. Thus, da7219 reports them as LINEOUT.
Adding the SND_JACK_LINEOUT type so that we don't fail to detect
any 3 pole jack type.
Also, changing
SND_JACK_HEADPHONE | SND_JACK_MICROPHONE -> SND_JACK_HEADSET
Signed-off-by: Akshu Agrawal <akshu.agrawal@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When running as a level 3 guest with no host provided sthyi support
sclp_ocf_cpc_name_copy() will only return zeroes. Zeroes are not a
valid group name, so let's not indicate that the group name field is
valid.
Also the group name is not dependent on stsi, let's not return based
on stsi before setting it.
Fixes: 95ca2cb579 ("KVM: s390: Add sthyi emulation")
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
With the new multi zcrypt device node support there came
in a code rework which broke the in-kernel api function
zcrypt_send_cprb(). This function is used by the pkey kernel
module and as an effect, transforming a secure key into a
protected key did not work any more.
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Since the core regulator code is treating GPIO descriptors as
nonexclusive, i.e. it assumes that the enable GPIO line may be
shared with several regulators, let's add the flag introduced
for fixing this problem on fixed regulators to all drivers
fetching GPIO descriptors to avoid possible regressions.
Reported-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
In some environments it is common that many hosts share the same lower half
of their IPv6 addresses (in particular ::1). As __xfrm6_addr_hash() and
__xfrm6_daddr_saddr_hash() calculate the hash only from the lower halves,
as much as 1/3 of the hosts ends up in one hashtable chain which harms the
performance.
Use complete IPv6 addresses when calculating the hashes. Rather than just
adding two more words to the xor, use jhash2() for consistency with
__xfrm6_pref_hash() and __xfrm6_dpref_spref_hash().
Signed-off-by: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
If we did some signal processing, we have to reload the pt_regs
tstate register because it's value may have changed.
In doing so we also have to extract the %pil value contained in there
anre load that into %l4.
This value is at bit 20 and thus needs to be shifted down before we
later write it into the %pil register.
Most of the time this is harmless as we are returning to userspace
and the %pil is zero for that case.
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit d595567dc4 (MD: fix invalid stored role for a disk) broke linear
hotadd. Let's only fix the role for disks in raid1/10.
Based on Guoqing's original patch.
Reported-by: kernel test robot <rong.a.chen@intel.com>
Cc: Gioh Kim <gi-oh.kim@profitbricks.com>
Cc: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Third set of patches for 4.20. Most notable is finalising ath10k
wcn3990 support, all components should be implemented now.
Major changes:
ath10k
* support NET_DETECT WoWLAN feature
* wcn3990 basic functionality now working after we got QMI support
mt76
* mt76x0e improvements (should be usable now)
* more mt76x0/mt76x2 unification work
brcmsmac
* fix a problem on AP mode with clients using power save mode
iwlwifi
* support for a new scan type: fast balance
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJbw1gWAAoJEG4XJFUm622b7IIH/18H4mJRAOG4PWDILroA6khv
xvZFWnBsJempGIfNqxUDJEaRggeFLerJG9rFGFUyAQ9qhaeJ67YJ6ySY6szvlpBX
VGZIn2EZ54OMWzRHOUAmQbuUsP5gW4MAM7wWOBGVf8mScFY2HIu7nUN+C2kDIqVU
091mbb1C3r3qS4uuo4VINSZG7EGA/+BkkbZWjIYAyEF4Ne0DX9LOSMKY2mKXnBbI
cgGQXBBGJdC425VZR1ErjRnQ3q+QfGCRfpKQu8aukhKvSG8Ew3WAEp8vyvqRS1LK
M25NMt7dLJWgvbhsQJtpeV4t6tgt6tAwI6u4mh8o2zTkpuNQLvrJYtr7Ufz/OFU=
=S4nY
-----END PGP SIGNATURE-----
Merge tag 'wireless-drivers-next-for-davem-2018-10-14' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next
Kalle Valo says:
====================
wireless-drivers-next patches for 4.20
Third set of patches for 4.20. Most notable is finalising ath10k
wcn3990 support, all components should be implemented now.
Major changes:
ath10k
* support NET_DETECT WoWLAN feature
* wcn3990 basic functionality now working after we got QMI support
mt76
* mt76x0e improvements (should be usable now)
* more mt76x0/mt76x2 unification work
brcmsmac
* fix a problem on AP mode with clients using power save mode
iwlwifi
* support for a new scan type: fast balance
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann says:
====================
pull-request: bpf 2018-10-14
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Fix xsk map update and delete operation to not call synchronize_net()
but to piggy back on SOCK_RCU_FREE for sockets instead as we are not
allowed to sleep under RCU, from Björn.
2) Do not change RLIMIT_MEMLOCK in reuseport_bpf selftest if the process
already has unlimited RLIMIT_MEMLOCK, from Eric.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
We can't modify cdo->capability as it is defined as a const.
Change the modification hack to just WARN_ON_ONCE() if we hit
any of the invalid combinations.
This fixes a regression for pcd, which doesn't work after the
constify patch.
Fixes: 853fe1bf75 ("cdrom: Make device operations read-only")
Tested-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We just allocated the queue and haven't even set it up yet,
hence we know that checking if ->mq_ops is NULL is always
going to be true.
In fact we do need to assign a lock to ->queue_lock always,
as we need it for the queue flags modifications.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Convert the driver to the modern blk-mq framework.
As byproduct we get rid of our open coded restart logic and let
blk-mq handle it.
Signed-off-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Straight forward conversion - instead of rewriting the internal buffer
retrieval logic, just replace the previous elevator peeking with an
internal list of requests.
Reviewed-by: "Ed L. Cashin" <ed.cashin@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When compiling the kernel with Clang, this warning appears even though
it is disabled for the whole kernel because this folder has its own set
of KBUILD_CFLAGS. It was disabled before the beginning of git history.
In file included from arch/x86/boot/compressed/kaslr.c:29:
In file included from arch/x86/boot/compressed/misc.h:21:
In file included from ./include/linux/elf.h:5:
In file included from ./arch/x86/include/asm/elf.h:77:
In file included from ./arch/x86/include/asm/vdso.h:11:
In file included from ./include/linux/mm_types.h:9:
In file included from ./include/linux/spinlock.h:88:
In file included from ./arch/x86/include/asm/spinlock.h:43:
In file included from ./arch/x86/include/asm/qrwlock.h:6:
./include/asm-generic/qrwlock.h:101:53: warning: passing 'u32 *' (aka
'unsigned int *') to parameter of type 'int *' converts between pointers
to integer types with different sign [-Wpointer-sign]
if (likely(atomic_try_cmpxchg_acquire(&lock->cnts, &cnts, _QW_LOCKED)))
^~~~~
./include/linux/compiler.h:76:40: note: expanded from macro 'likely'
# define likely(x) __builtin_expect(!!(x), 1)
^
./include/asm-generic/atomic-instrumented.h:69:66: note: passing
argument to parameter 'old' here
static __always_inline bool atomic_try_cmpxchg(atomic_t *v, int *old, int new)
^
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Link: https://lkml.kernel.org/r/20181013010713.6999-1-natechancellor@gmail.com
Clang warns that the declaration of jiffies in include/linux/jiffies.h
doesn't match the definition in arch/x86/time/kernel.c:
arch/x86/kernel/time.c:29:42: warning: section does not match previous declaration [-Wsection]
__visible volatile unsigned long jiffies __cacheline_aligned = INITIAL_JIFFIES;
^
./include/linux/cache.h:49:4: note: expanded from macro '__cacheline_aligned'
__section__(".data..cacheline_aligned")))
^
./include/linux/jiffies.h:81:31: note: previous attribute is here
extern unsigned long volatile __cacheline_aligned_in_smp __jiffy_arch_data jiffies;
^
./arch/x86/include/asm/cache.h:20:2: note: expanded from macro '__cacheline_aligned_in_smp'
__page_aligned_data
^
./include/linux/linkage.h:39:29: note: expanded from macro '__page_aligned_data'
#define __page_aligned_data __section(.data..page_aligned) __aligned(PAGE_SIZE)
^
./include/linux/compiler_attributes.h:233:56: note: expanded from macro '__section'
#define __section(S) __attribute__((__section__(#S)))
^
1 warning generated.
The declaration was changed in commit 7c30f352c8 ("jiffies.h: declare
jiffies and jiffies_64 with ____cacheline_aligned_in_smp") but wasn't
updated here. Make them match so Clang no longer warns.
Fixes: 7c30f352c8 ("jiffies.h: declare jiffies and jiffies_64 with ____cacheline_aligned_in_smp")
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181013005311.28617-1-natechancellor@gmail.com
Andi Kleen was just asking me about the NMI CR3 handling and why
we restore it unconditionally. I was *sure* we had documented it
well. We did not.
Add some documentation. We have common entry code where the CR3
value is stashed, but three places in two big code paths where we
restore it. I put bulk of the comments in this common path and
then refer to it from the other spots.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: luto@kernel.org
Cc: bp@alien8.de
Cc: "H. Peter Anvin" <hpa@zytor.come
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Link: https://lkml.kernel.org/r/20181012232118.3EAAE77B@viggo.jf.intel.com
Eric reported that a sequence count loop using this_cpu_read() got
optimized out. This is wrong, this_cpu_read() must imply READ_ONCE()
because the interface is IRQ-safe, therefore an interrupt can have
changed the per-cpu value.
Fixes: 7c3576d261 ("[PATCH] i386: Convert PDA into the percpu section")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Eric Dumazet <edumazet@google.com>
Cc: hpa@zytor.com
Cc: eric.dumazet@gmail.com
Cc: bp@alien8.de
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181011104019.748208519@infradead.org
Looking at the asm for native_sched_clock() I noticed we don't inline
enough. Mostly caused by sharing code with cyc2ns_read_begin(), which
we didn't used to do. So mark all that __force_inline to make it DTRT.
Fixes: 59eaef78bf ("x86/tsc: Remodel cyc2ns to use seqcount_latch()")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: hpa@zytor.com
Cc: eric.dumazet@gmail.com
Cc: bp@alien8.de
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181011104019.695196158@infradead.org
* mt76x0 fixes
* mt76x0e improvements (should be usable now)
* usb support improvements
* more mt76x0/mt76x2 unification work
* minor fix for aggregation + powersave clients
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2
Comment: GPGTools - http://gpgtools.org
iEYEABECAAYFAlvCEgoACgkQ130UHQKnbvU4RQCeKXCe/rgtrGQ0KXjcuZbXfYmn
2pAAoKLCTrI9QTCScMnTdyfHqiRCEBF8
=wK37
-----END PGP SIGNATURE-----
Merge tag 'mt76-for-kvalo-2018-10-13' of https://github.com/nbd168/wireless
mt76 patches for 4.20
* mt76x0 fixes
* mt76x0e improvements (should be usable now)
* usb support improvements
* more mt76x0/mt76x2 unification work
* minor fix for aggregation + powersave clients
The check on status is redundant as a status has to be zero at
the point it is being checked because of a previous check and return
path via label 'unlock'. Remove the redundant check and the deadcode
that can never be reached.
Detected by CoverityScan, CID#1471710 ("Logically dead code")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Add the result values specific to L2CAP LE credit based connections
and change the old result values wherever they were used.
Signed-off-by: Mallikarjun Phulari <mallikarjun.phulari@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
BCM43430 devices soldered onto the PCB (non-removable)
use an UART connection for bluetooth.
But also advertise btsdio support on their 3th sdio function.
Signed-off-by: Cho, Yu-Chen <acho@suse.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
The DMA API does its own zone decisions based on the coherent_dma_mask.
[ Note: as the driver doesn't set the DMA coherent mask, we can assume
the default 32bit DMA, hence it should be safe to drop the flag here
-- tiwai ]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Currently we limit the max addressable memory to 128TB. This patch increase the
limit to 2PB. We can have devices like nvdimm which adds memory above 512TB
limit.
We still don't support regular system ram above 512TB. One of the challenge with
that is the percpu allocator, that allocates per node memory and use the max
distance between them as the percpu offsets. This means with large gap in
address space ( system ram above 1PB) we will run out of vmalloc space to map
the percpu allocation.
In order to support addressable memory above 512TB, kernel should be able to
linear map this range. To do that with hash translation we now add 4 context
to kernel linear map region. Our per context addressable range is 512TB. We
still keep VMALLOC and VMEMMAP region to old size. SLB miss handlers is updated
to validate these limit.
We also limit this update to SPARSEMEM_VMEMMAP and SPARSEMEM_EXTREME
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
We will be adding get_kernel_context later. Update function name to indicate
this handle context allocation user space address.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This adds CONFIG_DEBUG_VM checks to ensure:
- The kernel stack is in the SLB after it's flushed and bolted.
- We don't insert an SLB for an address that is aleady in the SLB.
- The kernel SLB miss handler does not take an SLB miss.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
slb_flush_and_rebolt() is misleading, it is called in virtual mode, so
it can not possibly change the stack, so it should not be touching the
shadow area. And since vmalloc is no longer bolted, it should not
change any bolted mappings at all.
Change the name to slb_flush_and_restore_bolted(), and have it just
load the kernel stack from what's currently in the shadow SLB area.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
When switching processes, currently all user SLBEs are cleared, and a
few (exec_base, pc, and stack) are preloaded. In trivial testing with
small apps, this tends to miss the heap and low 256MB segments, and it
will also miss commonly accessed segments on large memory workloads.
Add a simple round-robin preload cache that just inserts the last SLB
miss into the head of the cache and preloads those at context switch
time. Every 256 context switches, the oldest entry is removed from the
cache to shrink the cache and require fewer slbmte if they are unused.
Much more could go into this, including into the SLB entry reclaim
side to track some LRU information etc, which would require a study of
large memory workloads. But this is a simple thing we can do now that
is an obvious win for common workloads.
With the full series, process switching speed on the context_switch
benchmark on POWER9/hash (with kernel speculation security masures
disabled) increases from 140K/s to 178K/s (27%).
POWER8 does not change much (within 1%), it's unclear why it does not
see a big gain like POWER9.
Booting to busybox init with 256MB segments has SLB misses go down
from 945 to 69, and with 1T segments 900 to 21. These could almost all
be eliminated by preloading a bit more carefully with ELF binary
loading.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This will be used by the SLB code in the next patch, but for now this
sets the slb_addr_limit to the correct size for 32-bit tasks.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Add 32-entry bitmaps to track the allocation status of the first 32
SLB entries, and whether they are user or kernel entries. These are
used to allocate free SLB entries first, before resorting to the round
robin allocator.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This patch moves SLB miss handlers completely to C, using the standard
exception handler macros to set up the stack and branch to C.
This can be done because the segment containing the kernel stack is
always bolted, so accessing it with relocation on will not cause an
SLB exception.
Arbitrary kernel memory must not be accessed when handling kernel
space SLB misses, so care should be taken there. However user SLB
misses can access any kernel memory, which can be used to move some
fields out of the paca (in later patches).
User SLB misses could quite easily reconcile IRQs and set up a first
class kernel environment and exit via ret_from_except, however that
doesn't seem to be necessary at the moment, so we only do that if a
bad fault is encountered.
[ Credit to Aneesh for bug fixes, error checks, and improvements to
bad address handling, etc ]
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Disallow tracing for all of slb.c for now.]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
