All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Michael Ellerman <mpe@ellerman.id.au>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 4.9 33/92] powerpc/io: Fix the IO workarounds code to work with Radix
Date: Thu, 29 Nov 2018 15:12:02 +0100	[thread overview]
Message-ID: <20181129140109.074225487@linuxfoundation.org> (raw)
In-Reply-To: <20181129140106.520639693@linuxfoundation.org>

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

[ Upstream commit 43c6494fa1499912c8177e71450c0279041152a6 ]

Back in 2006 Ben added some workarounds for a misbehaviour in the
Spider IO bridge used on early Cell machines, see commit
014da7ff47b5 ("[POWERPC] Cell "Spider" MMIO workarounds"). Later these
were made to be generic, ie. not tied specifically to Spider.

The code stashes a token in the high bits (59-48) of virtual addresses
used for IO (eg. returned from ioremap()). This works fine when using
the Hash MMU, but when we're using the Radix MMU the bits used for the
token overlap with some of the bits of the virtual address.

This is because the maximum virtual address is larger with Radix, up
to c00fffffffffffff, and in fact we use that high part of the address
range for ioremap(), see RADIX_KERN_IO_START.

As it happens the bits that are used overlap with the bits that
differentiate an IO address vs a linear map address. If the resulting
address lies outside the linear mapping we will crash (see below), if
not we just corrupt memory.

  virtio-pci 0000:00:00.0: Using 64-bit direct DMA at offset 800000000000000
  Unable to handle kernel paging request for data at address 0xc000000080000014
  ...
  CFAR: c000000000626b98 DAR: c000000080000014 DSISR: 42000000 IRQMASK: 0
  GPR00: c0000000006c54fc c00000003e523378 c0000000016de600 0000000000000000
  GPR04: c00c000080000014 0000000000000007 0fffffff000affff 0000000000000030
         ^^^^
  ...
  NIP [c000000000626c5c] .iowrite8+0xec/0x100
  LR [c0000000006c992c] .vp_reset+0x2c/0x90
  Call Trace:
    .pci_bus_read_config_dword+0xc4/0x120 (unreliable)
    .register_virtio_device+0x13c/0x1c0
    .virtio_pci_probe+0x148/0x1f0
    .local_pci_probe+0x68/0x140
    .pci_device_probe+0x164/0x220
    .really_probe+0x274/0x3b0
    .driver_probe_device+0x80/0x170
    .__driver_attach+0x14c/0x150
    .bus_for_each_dev+0xb8/0x130
    .driver_attach+0x34/0x50
    .bus_add_driver+0x178/0x2f0
    .driver_register+0x90/0x1a0
    .__pci_register_driver+0x6c/0x90
    .virtio_pci_driver_init+0x2c/0x40
    .do_one_initcall+0x64/0x280
    .kernel_init_freeable+0x36c/0x474
    .kernel_init+0x24/0x160
    .ret_from_kernel_thread+0x58/0x7c

This hasn't been a problem because CONFIG_PPC_IO_WORKAROUNDS which
enables this code is usually not enabled. It is only enabled when it's
selected by PPC_CELL_NATIVE which is only selected by
PPC_IBM_CELL_BLADE and that in turn depends on BIG_ENDIAN. So in order
to hit the bug you need to build a big endian kernel, with IBM Cell
Blade support enabled, as well as Radix MMU support, and then boot
that on Power9 using Radix MMU.

Still we can fix the bug, so let's do that. We simply use fewer bits
for the token, taking the union of the restrictions on the address
from both Hash and Radix, we end up with 8 bits we can use for the
token. The only user of the token is iowa_mem_find_bus() which only
supports 8 token values, so 8 bits is plenty for that.

Fixes: 566ca99af026 ("powerpc/mm/radix: Add dummy radix_enabled()")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/powerpc/include/asm/io.h | 20 +++++++-------------
 1 file changed, 7 insertions(+), 13 deletions(-)

diff --git a/arch/powerpc/include/asm/io.h b/arch/powerpc/include/asm/io.h
index f6fda8482f60..3b10f532c28e 100644
--- a/arch/powerpc/include/asm/io.h
+++ b/arch/powerpc/include/asm/io.h
@@ -333,19 +333,13 @@ extern void _memcpy_toio(volatile void __iomem *dest, const void *src,
  * their hooks, a bitfield is reserved for use by the platform near the
  * top of MMIO addresses (not PIO, those have to cope the hard way).
  *
- * This bit field is 12 bits and is at the top of the IO virtual
- * addresses PCI_IO_INDIRECT_TOKEN_MASK.
+ * The highest address in the kernel virtual space are:
  *
- * The kernel virtual space is thus:
+ *  d0003fffffffffff	# with Hash MMU
+ *  c00fffffffffffff	# with Radix MMU
  *
- *  0xD000000000000000		: vmalloc
- *  0xD000080000000000		: PCI PHB IO space
- *  0xD000080080000000		: ioremap
- *  0xD0000fffffffffff		: end of ioremap region
- *
- * Since the top 4 bits are reserved as the region ID, we use thus
- * the next 12 bits and keep 4 bits available for the future if the
- * virtual address space is ever to be extended.
+ * The top 4 bits are reserved as the region ID on hash, leaving us 8 bits
+ * that can be used for the field.
  *
  * The direct IO mapping operations will then mask off those bits
  * before doing the actual access, though that only happen when
@@ -357,8 +351,8 @@ extern void _memcpy_toio(volatile void __iomem *dest, const void *src,
  */
 
 #ifdef CONFIG_PPC_INDIRECT_MMIO
-#define PCI_IO_IND_TOKEN_MASK	0x0fff000000000000ul
-#define PCI_IO_IND_TOKEN_SHIFT	48
+#define PCI_IO_IND_TOKEN_SHIFT	52
+#define PCI_IO_IND_TOKEN_MASK	(0xfful << PCI_IO_IND_TOKEN_SHIFT)
 #define PCI_FIX_ADDR(addr)						\
 	((PCI_IO_ADDR)(((unsigned long)(addr)) & ~PCI_IO_IND_TOKEN_MASK))
 #define PCI_GET_ADDR_TOKEN(addr)					\
-- 
2.17.1




  parent reply	other threads:[~2018-11-29 14:21 UTC|newest]

Thread overview: 103+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-29 14:11 [PATCH 4.9 00/92] 4.9.142-stable review Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 01/92] usb: core: Fix hub port connection events lost Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 02/92] usb: dwc3: core: Clean up ULPI device Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 03/92] usb: xhci: fix timeout for transition from RExit to U0 Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 04/92] MAINTAINERS: Add Sasha as a stable branch maintainer Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 05/92] gpio: dont free unallocated ida on gpiochip_add_data_with_key() error path Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 06/92] iwlwifi: mvm: support sta_statistics() even on older firmware Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 07/92] iwlwifi: mvm: fix regulatory domain update when the firmware starts Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 08/92] brcmfmac: fix reporting support for 160 MHz channels Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 09/92] tools/power/cpupower: fix compilation with STATIC=true Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 10/92] v9fs_dir_readdir: fix double-free on p9stat_read error Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 11/92] selinux: Add __GFP_NOWARN to allocation at str_read() Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 12/92] bfs: add sanity check at bfs_fill_super() Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 13/92] sctp: clear the transport of some out_chunk_list chunks in sctp_assoc_rm_peer Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 14/92] gfs2: Dont leave s_fs_info pointing to freed memory in init_sbd Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 15/92] llc: do not use sk_eat_skb() Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 16/92] mm: dont warn about large allocations for slab Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 17/92] drm/ast: change resolution may cause screen blurred Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 18/92] drm/ast: fixed cursor may disappear sometimes Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 19/92] drm/ast: Remove existing framebuffers before loading driver Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 20/92] can: dev: can_get_echo_skb(): factor out non sending code to __can_get_echo_skb() Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 21/92] can: dev: __can_get_echo_skb(): replace struct can_frame by canfd_frame to access frame length Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 22/92] can: dev: __can_get_echo_skb(): Dont crash the kernel if can_priv::echo_skb is accessed out of bounds Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 23/92] can: dev: __can_get_echo_skb(): print error message, if trying to echo non existing skb Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 24/92] ACPICA: AML interpreter: add region addresses in global list during initialization Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 25/92] IB/core: Fix for core panic Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 26/92] IB/hfi1: Eliminate races in the SDMA send error path Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 27/92] usb: xhci: Prevent bus suspend if a port connect change or polling state is detected Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 28/92] pinctrl: meson: fix pinconf bias disable Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 29/92] KVM: PPC: Move and undef TRACE_INCLUDE_PATH/FILE Greg Kroah-Hartman
2018-11-29 14:11 ` [PATCH 4.9 30/92] cpufreq: imx6q: add return value check for voltage scale Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 31/92] rtc: pcf2127: fix a kmemleak caused in pcf2127_i2c_gather_write Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 32/92] floppy: fix race condition in __floppy_read_block_0() Greg Kroah-Hartman
2018-11-29 14:12 ` Greg Kroah-Hartman [this message]
2018-11-29 14:12 ` [PATCH 4.9 34/92] perf/x86/intel/uncore: Add more IMC PCI IDs for KabyLake and CoffeeLake CPUs Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 35/92] ARM: make lookup_processor_type() non-__init Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 36/92] SUNRPC: Fix a bogus get/put in generic_key_to_expire() Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 37/92] kdb: Use strscpy with destination buffer size Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 38/92] powerpc/numa: Suppress "VPHN is not supported" messages Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 39/92] efi/arm: Revert deferred unmap of early memmap mapping Greg Kroah-Hartman
2018-11-29 14:28   ` Ard Biesheuvel
2018-11-29 15:02     ` Greg Kroah-Hartman
2018-11-29 15:03       ` Ard Biesheuvel
2018-11-29 14:12 ` [PATCH 4.9 40/92] tmpfs: make lseek(SEEK_DATA/SEK_HOLE) return ENXIO with a negative offset Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 41/92] of: add helper to lookup compatible child node Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 42/92] NFC: nfcmrvl_uart: fix OF child-node lookup Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 43/92] net: bcmgenet: " Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 44/92] arm64: remove no-op -p linker flag Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 45/92] ath10k: fix kernel panic due to race in accessing arvif list Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 46/92] Input: xpad - add product ID for Xbox One S pad Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 47/92] Input: xpad - fix Xbox One rumble stopping after 2.5 secs Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 48/92] Input: xpad - correctly sort vendor ids Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 49/92] Input: xpad - move reporting xbox one home button to common function Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 50/92] Input: xpad - simplify error condition in init_output Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 51/92] Input: xpad - dont depend on endpoint order Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 52/92] Input: xpad - fix stuck mode button on Xbox One S pad Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 53/92] Input: xpad - restore LED state after device resume Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 54/92] Input: xpad - support some quirky Xbox One pads Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 55/92] Input: xpad - sort supported devices by USB ID Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 56/92] Input: xpad - sync supported devices with xboxdrv Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 57/92] Input: xpad - add USB IDs for Mad Catz Brawlstick and Razer Sabertooth Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 58/92] Input: xpad - sync supported devices with 360Controller Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 59/92] Input: xpad - sync supported devices with XBCD Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 60/92] Input: xpad - constify usb_device_id Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 61/92] Input: xpad - fix PowerA init quirk for some gamepad models Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 62/92] Input: xpad - validate USB endpoint type during probe Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 63/92] Input: xpad - add support for PDP Xbox One controllers Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 64/92] Input: xpad - add PDP device id 0x02a4 Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 65/92] Input: xpad - fix some coding style issues Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 66/92] Input: xpad - avoid using __set_bit() for capabilities Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 67/92] Input: xpad - add GPD Win 2 Controller USB IDs Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 68/92] Input: xpad - fix GPD Win 2 controller name Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 69/92] Input: xpad - add support for Xbox1 PDP Camo series gamepad Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 70/92] cw1200: Dont leak memory if krealloc failes Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 71/92] mwifiex: prevent register accesses after host is sleeping Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 72/92] mwifiex: report error to PCIe for suspend failure Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 73/92] mwifiex: Fix NULL pointer dereference in skb_dequeue() Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 74/92] mwifiex: fix p2p device doesnt find in scan problem Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 75/92] scsi: ufs: fix bugs related to null pointer access and array size Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 76/92] scsi: ufshcd: Fix race between clk scaling and ungate work Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 77/92] scsi: ufs: fix race between clock gating and devfreq scaling work Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 78/92] scsi: ufshcd: release resources if probe fails Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 79/92] include/linux/pfn_t.h: force ~ to be parsed as an unary operator Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 80/92] tty: wipe buffer Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 81/92] tty: wipe buffer if not echoing data Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 82/92] usb: xhci: fix uninitialized completion when USB3 port got wrong status Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 83/92] sched/core: Allow __sched_setscheduler() in interrupts when PI is not used Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 84/92] namei: allow restricted O_CREAT of FIFOs and regular files Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 85/92] lan78xx: Read MAC address from DT if present Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 86/92] s390/mm: Check for valid vma before zapping in gmap_discard Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 87/92] net: ieee802154: 6lowpan: fix frag reassembly Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 88/92] Revert "evm: Translate user/group ids relative to s_user_ns when computing HMAC" Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 89/92] ima: always measure and audit files in policy Greg Kroah-Hartman
2018-11-29 14:12 ` [PATCH 4.9 90/92] EVM: Add support for portable signature format Greg Kroah-Hartman
2018-11-29 14:13 ` [PATCH 4.9 91/92] ima: re-introduce own integrity cache lock Greg Kroah-Hartman
2018-11-29 14:13 ` [PATCH 4.9 92/92] ima: re-initialize iint->atomic_flags Greg Kroah-Hartman
2018-11-29 19:51 ` [PATCH 4.9 00/92] 4.9.142-stable review kernelci.org bot
2018-11-29 20:29 ` shuah
2018-11-30  7:08 ` Naresh Kamboju
2018-11-30  9:10 ` Jon Hunter
2018-11-30  9:10   ` Jon Hunter
2018-11-30 10:36   ` Greg Kroah-Hartman
2018-11-30 22:28 ` Guenter Roeck

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20181129140109.074225487@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mpe@ellerman.id.au \
    --cc=sashal@kernel.org \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.