| Age | Commit message (Collapse) | Author |
|---|
|
On a system with a repeater on the system board to support gen2 hotplug,
we found that when an ExpressModule is removed from some slots,
/var/log/messages will be full of "card present/not present" warnings.
It turns out the root complex is continually trying to train the link to
the repeater because the repeater has not been reset.
This patch will disable the link at removal time to allow the repeater
to be reset properly. This also prevents a potential AER message at
removal time.
Also, when testing hotplug on a system under development, we found if we
boot the system without an EM installed, and later hot-add an EM, it
does not work with Linux, but another OS is ok. The root cause is that
BIOS left link disabled when slot was empty at boot time, and other OS
is modifying the link disable bit in link ctrl during power on/off.
So we should do the same thing to disable/enable link during power off/on.
-v2: check link DLLA bit instead of 100ms waiting.
Separate link disable/enable functions to another patch.
Signed-off-by: Yinghai Lu <yinghai.lu@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Will use it during power off/on of slots
Signed-off-by: Yinghai Lu <yinghai.lu@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Will use it for link disable status checking.
Signed-off-by: Yinghai Lu <yinghai.lu@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
A few changes:
- remove the 'inline' and let the complier decide
- return a bool to indicate whether the link was active
- add a debug message to indicate link state when it beocmes active
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
During reviewing
| PCI: pciehp: wait 1000 ms before Link Training check
Linus said:
>...
> That's a *long* time, and it's irritating to the user. It makes the
> user think "the machine is slow".
>...
> And quite frankly, an unconditional one-second delay here seems bad.
>Two seconds was unacceptable, one second is just bad.
Try to access the pci conf of a pci device that is supposed to show up
in 1s. If we can read back a valid vendor/device id, we can return
early.
Related discussion could be found:
https://lkml.org/lkml/2011/12/6/339
-v2: seperate code to pci_bus_read_dev_vendor_id() from pci_scan_device()
and reuse it from pciehp code. Suggested by Matthew Wilcox.
-v3: According to Kenj, don't use array in stack, and don't wait too long
for crs, also return fail status if not found.
Also separate pci_bus_dev_read_vendor_id() change to another patch.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
We can reuse it for pciehp probing.
-v2: according to Kenji, fix crs timeout checking, and export the function
for later use when pciehp is compiled as a module.
Suggested-by: Matthew Wilcox <matthew@wil.cx>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
When hot removing a pci express module that has a pcie switch and supports
SRIOV, we got:
[ 5918.610127] pciehp 0000:80:02.2:pcie04: pcie_isr: intr_loc 1
[ 5918.615779] pciehp 0000:80:02.2:pcie04: Attention button interrupt received
[ 5918.622730] pciehp 0000:80:02.2:pcie04: Button pressed on Slot(3)
[ 5918.629002] pciehp 0000:80:02.2:pcie04: pciehp_get_power_status: SLOTCTRL a8 value read 1f9
[ 5918.637416] pciehp 0000:80:02.2:pcie04: PCI slot #3 - powering off due to button press.
[ 5918.647125] pciehp 0000:80:02.2:pcie04: pcie_isr: intr_loc 10
[ 5918.653039] pciehp 0000:80:02.2:pcie04: pciehp_green_led_blink: SLOTCTRL a8 write cmd 200
[ 5918.661229] pciehp 0000:80:02.2:pcie04: pciehp_set_attention_status: SLOTCTRL a8 write cmd c0
[ 5924.667627] pciehp 0000:80:02.2:pcie04: Disabling domain:bus:device=0000:b0:00
[ 5924.674909] pciehp 0000:80:02.2:pcie04: pciehp_get_power_status: SLOTCTRL a8 value read 2f9
[ 5924.683262] pciehp 0000:80:02.2:pcie04: pciehp_unconfigure_device: domain:bus:dev = 0000:b0:00
[ 5924.693976] libfcoe_device_notification: NETDEV_UNREGISTER eth6
[ 5924.764979] libfcoe_device_notification: NETDEV_UNREGISTER eth14
[ 5924.873539] libfcoe_device_notification: NETDEV_UNREGISTER eth15
[ 5924.995209] libfcoe_device_notification: NETDEV_UNREGISTER eth16
[ 5926.114407] sxge 0000:b2:00.0: PCI INT A disabled
[ 5926.119342] BUG: unable to handle kernel NULL pointer dereference at (null)
[ 5926.127189] IP: [<ffffffff81353a3b>] pci_stop_bus_device+0x33/0x83
[ 5926.133377] PGD 0
[ 5926.135402] Oops: 0000 [#1] SMP
[ 5926.138659] CPU 2
[ 5926.140499] Modules linked in:
...
[ 5926.143754]
[ 5926.275823] Call Trace:
[ 5926.278267] [<ffffffff81353a38>] pci_stop_bus_device+0x30/0x83
[ 5926.284180] [<ffffffff81353af4>] pci_remove_bus_device+0x1a/0xba
[ 5926.290264] [<ffffffff81366311>] pciehp_unconfigure_device+0x110/0x17b
[ 5926.296866] [<ffffffff81365dd9>] ? pciehp_disable_slot+0x188/0x188
[ 5926.303123] [<ffffffff81365d6f>] pciehp_disable_slot+0x11e/0x188
[ 5926.309206] [<ffffffff81365e68>] pciehp_power_thread+0x8f/0xe0
...
+-[0000:80]-+-00.0-[81-8f]--
| +-01.0-[90-9f]--
| +-02.0-[a0-af]--
| +-02.2-[b0-bf]----00.0-[b1-b3]--+-02.0-[b2]--+-00.0 Device
| | | +-00.1 Device
| | | +-00.2 Device
| | | \-00.3 Device
| | \-03.0-[b3]--+-00.0 Device
| | +-00.1 Device
| | +-00.2 Device
| | \-00.3 Device
root complex: 80:02.2
pci express modules: have pcie switch and are listed as b0:00.0, b1:02.0 and b1:03.0.
end devices are b2:00.0 and b3.00.0.
VFs are: b2:00.1,... b2:00.3, and b3:00.1,...,b3:00.3
Root cause: when doing pci_stop_bus_device() with phys fn, it will stop
virt fn and remove the fn, so
list_for_each_safe(l, n, &bus->devices)
will have problem to refer freed n that is pointed to vf entry.
Solution is just replacing list_for_each_safe() with
list_for_each_prev_safe(). This will make sure we can get valid n pointer
to PF instead of the freed VF pointer (because newly added devices are
inserted to the bus->devices list tail).
During reviewing the patch, Bjorn said:
| The PCI hot-remove path calls pci_stop_bus_devices() via
| pci_remove_bus_device().
|
| pci_stop_bus_devices() traverses the bus->devices list (point A below),
| stopping each device in turn, which calls the driver remove() method. When
| the device is an SR-IOV PF, the driver calls pci_disable_sriov(), which
| also uses pci_remove_bus_device() to remove the VF devices from the
| bus->devices list (point B).
|
| pci_remove_bus_device
| pci_stop_bus_device
| pci_stop_bus_devices(subordinate)
| list_for_each(bus->devices) <-- A
| pci_stop_bus_device(PF)
| ...
| driver->remove
| pci_disable_sriov
| ...
| pci_remove_bus_device(VF)
| <remove from bus_list> <-- B
|
| At B, we're changing the same list we're iterating through at A, so when
| the driver remove() method returns, the pci_stop_bus_devices() iterator has
| a pointer to a list entry that has already been freed.
Discussion thread can be found : https://lkml.org/lkml/2011/10/15/141
https://lkml.org/lkml/2012/1/23/360
-v5: According to Linus to make remove more robust, Change to
list_for_each_prev_safe instead. That is more reasonable, because
those devices are added to tail of the list before.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Only one user; just use add_to_list instead.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
For use in debugging resource reallocation.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
After merging struct pci_dev_resource_x and pci_dev_resource,
We can use a function instead of macro now.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Linus says don't use dev_res_x because it doesn't communicate anything
about usage. Rename them to add_res or fail_res etc according to
context.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
pci_dev_resource_x is a superset of pci_dev_resource and they're just
temp structs used during resource reallocation.
pci_dev_resource usage is quite limted.
So just use pci_dev_resource_x, and rename it as new pci_dev_resource.
-v2: According to Linus, Separate free_list change to another patch
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
So we can use helper functions for generic list. This makes the
resource re-allocation code much more readable.
-v2: Use list_add_tail instead of adding list_insert_before, Pointed out
by Linus.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
No user outside of setup-bus.c now. Later patches will convert
resource_list to a regular list.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
This allows us to move the definition of struct resource_list to
setup_bus.c and later convert resource_list to a regular list.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
On a system with devices that support SRIOV connected to a pcie switch
to pcie root port:
+-[0000:80]-+-00.0-[81-8f]--
| +-01.0-[90-9f]--
| +-02.0-[a0-af]----00.0-[a1-a3]--+-02.0-[a2]--+-00.0 Oracle Corporation Device 207a
| | \-03.0-[a3]--+-00.0 Oracle Corporation Device 207a
| +-02.2-[b0-bf]----00.0-[b1-b3]--+-02.0-[b2]--+-00.0 Oracle Corporation Device 207a
| | \-03.0-[b3]--+-00.0 Oracle Corporation Device 207a
When the BIOS does not assign resources for SRIOV BARs, kernel pci
reallocation only goes up one bridge and then gives up, failing to to
get resources for all sSRIOV BARs, even though the range is large enough
in the peer root bus.
Specifically, only the bridge at the a1:02.0 level has its resources
cleared and reallocated. The kernel does not go up to clear the bridge
at the 80:02.0 level.
To make it go to upper levels, during retry, we need to treat "good to have"
resources as "must have".
Only on the last try will we treat good to have resources as optional.
At that time, parent bridge resources will already have been released so
we'll have a chance to get everything assigned with must_have plus
good_to_have for all child devices.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
This allows us to allocate resources to hotplug bridges during
remove/rescan.
We need to move the function to setup-bus.c so it can use
__pci_bus_size_bridges and __pci_bus_assign_resources directly to take
the add_list resource tracking list.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Current rescan will not touch bridge MMIO and IO.
Try to reuse pci_assign_unassigned_bridge_resources(bridge) to update bridge
resources, if child devices need more resources.
Only do that for bridges whose children are all removed already; i.e. don't
release resources that could already be in use by drivers on child devices.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
We need add size for hot plug path when pluging in hotplug chassis
without cards.
-v2: change descriptions. make it applicable after "pci: Check bridge
resources after resource allocation."
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
We found reassignment can not find a range for one resource, even if the
total available range is large enough.
bridge b1:02.0 will need 2M+3M
bridge b1:03.0 will need 2M+3M
so bridge b0:00.0 will get assigned: 4M : [f8000000-f83fffff]
later is reassigned to 10M : [f8000000-f9ffffff]
b1:02.0 is assigned to 2M : [f8000000-f81fffff]
b1:03.0 is assigned to 2M : [f8200000-f83fffff]
After that b1:03.0 get chance to be reassigned to [f8200000-f86fffff],
but b1:02.0 will not have chance to expand, because b1:03.0 is using in
middle one.
[ 187.911401] pci 0000:b1:02.0: bridge window [mem 0x00100000-0x002fffff] to [bus b2-b2] add_size 300000
[ 187.920764] pci 0000:b1:03.0: bridge window [mem 0x00100000-0x002fffff] to [bus b3-b3] add_size 300000
[ 187.930129] pci 0000:b1:02.0: [mem 0x00100000-0x002fffff] get_res_add_size add_size 300000
[ 187.938500] pci 0000:b1:03.0: [mem 0x00100000-0x002fffff] get_res_add_size add_size 300000
[ 187.946857] pci 0000:b0:00.0: bridge window [mem 0x00100000-0x004fffff] to [bus b1-b3] add_size 600000
[ 187.956206] pci 0000:b0:00.0: BAR 14: assigned [mem 0xf8000000-0xf83fffff]
[ 187.963102] pci 0000:b0:00.0: BAR 15: assigned [mem 0xf5000000-0xf51fffff pref]
[ 187.970434] pci 0000:b0:00.0: BAR 14: reassigned [mem 0xf8000000-0xf89fffff]
[ 187.977497] pci 0000:b1:02.0: BAR 14: assigned [mem 0xf8000000-0xf81fffff]
[ 187.984383] pci 0000:b1:02.0: BAR 15: assigned [mem 0xf5000000-0xf50fffff pref]
[ 187.991695] pci 0000:b1:03.0: BAR 14: assigned [mem 0xf8200000-0xf83fffff]
[ 187.998576] pci 0000:b1:03.0: BAR 15: assigned [mem 0xf5100000-0xf51fffff pref]
[ 188.005888] pci 0000:b1:03.0: BAR 14: reassigned [mem 0xf8200000-0xf86fffff]
[ 188.012939] pci 0000:b1:02.0: BAR 14: can't assign mem (size 0x200000)
[ 188.019471] pci 0000:b1:02.0: failed to add 300000 to res=[mem 0xf8000000-0xf81fffff]
[ 188.027326] pci 0000:b2:00.0: reg 184: [mem 0x00000000-0x00003fff 64bit]
[ 188.034071] pci 0000:b2:00.0: reg 18c: [mem 0x00000000-0x000fffff 64bit]
[ 188.040795] pci 0000:b2:00.0: BAR 2: assigned [mem 0xf8000000-0xf80fffff 64bit]
[ 188.048119] pci 0000:b2:00.0: BAR 2: set to [mem 0xf8000000-0xf80fffff 64bit] (PCI address [0xf8000000-0xf80fffff])
[ 188.058550] pci 0000:b2:00.0: BAR 6: assigned [mem 0xf5000000-0xf50fffff pref]
[ 188.065802] pci 0000:b2:00.0: BAR 0: assigned [mem 0xf8100000-0xf8103fff 64bit]
[ 188.073125] pci 0000:b2:00.0: BAR 0: set to [mem 0xf8100000-0xf8103fff 64bit] (PCI address [0xf8100000-0xf8103fff])
[ 188.083596] pci 0000:b2:00.0: reg 18c: [mem 0x00000000-0x000fffff 64bit]
[ 188.090310] pci 0000:b2:00.0: BAR 9: can't assign mem (size 0x300000)
[ 188.096773] pci 0000:b2:00.0: reg 184: [mem 0x00000000-0x00003fff 64bit]
[ 188.103479] pci 0000:b2:00.0: BAR 7: assigned [mem 0xf8104000-0xf810ffff 64bit]
[ 188.110801] pci 0000:b2:00.0: BAR 7: set to [mem 0xf8104000-0xf810ffff 64bit] (PCI address [0xf8104000-0xf810ffff])
[ 188.121256] pci 0000:b1:02.0: PCI bridge to [bus b2-b2]
[ 188.126512] pci 0000:b1:02.0: bridge window [mem 0xf8000000-0xf81fffff]
[ 188.133328] pci 0000:b1:02.0: bridge window [mem 0xf5000000-0xf50fffff pref]
[ 188.140608] pci 0000:b3:00.0: reg 184: [mem 0x00000000-0x00003fff 64bit]
[ 188.147341] pci 0000:b3:00.0: reg 18c: [mem 0x00000000-0x000fffff 64bit]
[ 188.154076] pci 0000:b3:00.0: BAR 2: assigned [mem 0xf8200000-0xf82fffff 64bit]
[ 188.161417] pci 0000:b3:00.0: BAR 2: set to [mem 0xf8200000-0xf82fffff 64bit] (PCI address [0xf8200000-0xf82fffff])
[ 188.171865] pci 0000:b3:00.0: BAR 6: assigned [mem 0xf5100000-0xf51fffff pref]
[ 188.179090] pci 0000:b3:00.0: BAR 0: assigned [mem 0xf8300000-0xf8303fff 64bit]
[ 188.186431] pci 0000:b3:00.0: BAR 0: set to [mem 0xf8300000-0xf8303fff 64bit] (PCI address [0xf8300000-0xf8303fff])
[ 188.196884] pci 0000:b3:00.0: reg 18c: [mem 0x00000000-0x000fffff 64bit]
[ 188.203591] pci 0000:b3:00.0: BAR 9: assigned [mem 0xf8400000-0xf86fffff 64bit]
[ 188.210909] pci 0000:b3:00.0: BAR 9: set to [mem 0xf8400000-0xf86fffff 64bit] (PCI address [0xf8400000-0xf86fffff])
[ 188.221379] pci 0000:b3:00.0: reg 184: [mem 0x00000000-0x00003fff 64bit]
[ 188.228089] pci 0000:b3:00.0: BAR 7: assigned [mem 0xf8304000-0xf830ffff 64bit]
[ 188.235407] pci 0000:b3:00.0: BAR 7: set to [mem 0xf8304000-0xf830ffff 64bit] (PCI address [0xf8304000-0xf830ffff])
[ 188.245843] pci 0000:b1:03.0: PCI bridge to [bus b3-b3]
[ 188.251107] pci 0000:b1:03.0: bridge window [mem 0xf8200000-0xf86fffff]
[ 188.257922] pci 0000:b1:03.0: bridge window [mem 0xf5100000-0xf51fffff pref]
[ 188.265180] pci 0000:b0:00.0: PCI bridge to [bus b1-b3]
[ 188.270443] pci 0000:b0:00.0: bridge window [mem 0xf8000000-0xf89fffff]
[ 188.277250] pci 0000:b0:00.0: bridge window [mem 0xf5000000-0xf51fffff pref]
[ 188.284512] pcieport 0000:80:02.2: PCI bridge to [bus b0-bf]
[ 188.290184] pcieport 0000:80:02.2: bridge window [io 0xa000-0xbfff]
[ 188.296735] pcieport 0000:80:02.2: bridge window [mem 0xf8000000-0xf8ffffff]
[ 188.303963] pcieport 0000:80:02.2: bridge window [mem 0xf5000000-0xf5ffffff 64bit pref]
Thus b2:00.0 BAR 9 does not get assigned...
root cause:
b1:02.0 can not be added more range, because b1:03.0 is just after it;
no space between the required ranges.
Solution:
Try to assign required + optional all together at first, and if that
fails, try again with just the required resources.
-v2: seperate add_to_list change() to another patch according to Jesse.
seperate get_res_add_size() moving to another patch according to Jesse.
add !realloc_head->next check if the list is empty to bail early
according to Jesse.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Need to call it from __assign_resources_sorted() later and we'd like to
avoid a forward declaraion.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Will be used for resource_list_x duplication when trying
requested+optional at first.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
During debug of one SRIOV enabled hotplug device, we found found that
add_size is not passed properly.
The device has devices under two level bridges:
+-[0000:80]-+-00.0-[81-8f]--
| +-01.0-[90-9f]--
| +-02.0-[a0-af]----00.0-[a1-a3]--+-02.0-[a2]--+-00.0 Oracle Corporation Device
| | \-03.0-[a3]--+-00.0 Oracle Corporation Device
Which means later the parent bridge will not try to add a big enough range:
[ 557.455077] pci 0000:a0:00.0: BAR 14: assigned [mem 0xf9000000-0xf93fffff]
[ 557.461974] pci 0000:a0:00.0: BAR 15: assigned [mem 0xf6000000-0xf61fffff pref]
[ 557.469340] pci 0000:a1:02.0: BAR 14: assigned [mem 0xf9000000-0xf91fffff]
[ 557.476231] pci 0000:a1:02.0: BAR 15: assigned [mem 0xf6000000-0xf60fffff pref]
[ 557.483582] pci 0000:a1:03.0: BAR 14: assigned [mem 0xf9200000-0xf93fffff]
[ 557.490468] pci 0000:a1:03.0: BAR 15: assigned [mem 0xf6100000-0xf61fffff pref]
[ 557.497833] pci 0000:a1:03.0: BAR 14: can't assign mem (size 0x200000)
[ 557.504378] pci 0000:a1:03.0: failed to add optional resources res=[mem 0xf9200000-0xf93fffff]
[ 557.513026] pci 0000:a1:02.0: BAR 14: can't assign mem (size 0x200000)
[ 557.519578] pci 0000:a1:02.0: failed to add optional resources res=[mem 0xf9000000-0xf91fffff]
It turns out we did not calculate size1 properly.
static resource_size_t calculate_memsize(resource_size_t size,
resource_size_t min_size,
resource_size_t size1,
resource_size_t old_size,
resource_size_t align)
{
if (size < min_size)
size = min_size;
if (old_size == 1 )
old_size = 0;
if (size < old_size)
size = old_size;
size = ALIGN(size + size1, align);
return size;
}
We should not pass add_size with min_size in calculate_memsize since
that will make add_size not contribute final add_size.
So just pass add_size with size1 to calculate_memsize().
With this change, we should have chance to remove extra addon in
pci_reassign_resource.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Correct spelling "resouce" to "resource" in
dricers/pci/setup-res.c
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
We use the __pci_reset_function_locked to perform the action.
Also on attaching ("bind") and detaching ("unbind") we save and
restore the configuration states. When the device is disconnected
from a guest we use the "pci_reset_function" to also reset the
device before being passed to another guest.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
The use case of this is when a driver wants to call FLR when a device
is attached to it using the SysFS "bind" or "unbind" functionality.
The call chain when a user does "bind" looks as so:
echo "0000:01.07.0" > /sys/bus/pci/drivers/XXXX/bind
and ends up calling:
driver_bind:
device_lock(dev); <=== TAKES LOCK
XXXX_probe:
.. pci_enable_device()
...__pci_reset_function(), which calls
pci_dev_reset(dev, 0):
if (!0) {
device_lock(dev) <==== DEADLOCK
The __pci_reset_function_locked function allows the the drivers
'probe' function to call the "pci_reset_function" while still holding
the driver mutex lock.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Add missing iounmap in error handling code, in a case where the function
already preforms iounmap on some other execution path.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
@@
expression e;
statement S,S1;
int ret;
@@
e = \(ioremap\|ioremap_nocache\)(...)
... when != iounmap(e)
if (<+...e...+>) S
... when any
when != iounmap(e)
*if (...)
{ ... when != iounmap(e)
return ...; }
... when any
iounmap(e);
// </smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Boot up a KVM guest, and hotplug multifunction
devices(func1,func2,func0,func3) to guest.
for i in 1 2 0 3;do
qemu-img create /tmp/resize$i.qcow2 1G -f qcow2
(qemu) drive_add 0x11.$i id=drv11$i,if=none,file=/tmp/resize$i.qcow2
(qemu) device_add virtio-blk-pci,id=dev11$i,drive=drv11$i,addr=0x11.$i,multifunction=on
done
In linux kernel, when func0 of the slot is hot-added, the whole
slot will be marked as 'enabled', then driver will ignore other new
hotadded funcs.
But in Win7 & WinXP, we can continaully add other funcs after adding
func0, all funcs will be added in guest.
drivers/pci/hotplug/acpiphp_glue.c:
static int acpiphp_check_bridge(struct acpiphp_bridge *bridge)
{
....
for (slot = bridge->slots; slot; slot = slot->next) {
if (slot->flags & SLOT_ENABLED) {
acpiphp_disable_slot()
else
acpiphp_enable_slot()
.... |
} v
enable_device()
|
v
//only don't enable slot if func0 is not added
list_for_each_entry(func, &slot->funcs, sibling) {
...
}
slot->flags |= SLOT_ENABLED; //mark slot to 'enabled'
This patch just make pci driver can continaully add funcs after adding
func 0. Only mark slot to 'enabled' when all funcs are added.
For pci multifunction hotplug, we can add functions one by one(func 0 is
necessary), and all functions will be removed in one time.
Signed-off-by: Amos Kong <akong@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
This patch converts the underlying maintenance aspects of FW-assigned
BIOS BAR values from a statically allocated array within struct pci_dev
to a list of temporary, stand alone, entries.
Signed-off-by: Myron Stowe <myron.stowe@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
pci_revert_fw_address() is used to reinstate a PCI device's original
FW-assigned BIOS BAR value(s) if normal resource assignment fails.
When attempting to reinstate an address, the point within the resource
tree from which to attempt the new resource request should be the parent
resource corresponding to the device, not the base of the resource tree
(ioport_resource or iomem_resource). For PCI devices this would
typically be the resource corresponding to the upstream PCI host bridge
or P2P bridge aperture.
This patch sets the point within the resource tree to attempt a new
resource assignment request to the PCI device's parent resource and only
if that fails does it fall back to the base ioport_resource or
iomem_resource.
Signed-off-by: Myron Stowe <myron.stowe@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
During test busn_res allocation with cardbus, found pci card removal is not
working anymore, and it turns out it is broken by:
|commit 79cc9601c3e42b4f0650fe7e69132ebce7ab48f9
|Date: Tue Nov 22 21:06:53 2011 -0800
|
| PCI: Only call pci_stop_bus_device() one time for child devices at remove
The above changed the behavior of pci_remove_behind_bridge that
yenta_cardbus depended on. So restore the old behavoir of
pci_remove_behind_bridge (which requires stopping and removing of all
devices) by:
1. rename pci_remove_behind_bridge to __pci_remove_behind_bridge, and let
__pci_remove_bus_device() call it instead.
2. add pci_stop_behind_bridge that will stop devices behind a bridge
3. add back pci_remove_behind_bridge that will stop and remove devices
under bridge.
-v2: update commit description a little bit.
Tested-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
For an SRIOV device, PCI_SRIOV_SYS_PGSIZE should be set before
the PCI_SRIOV_BAR are queried. The sys pagesize defaults to 4k,
so this change is required on powerpc box with 64k base page size.
This is a regression caused due to moving SRIOV init to sriov_enable().
| commit afd24ece5c76af87f6fc477f2747b83a764f161c
| Author: Ram Pai <linuxram@us.ibm.com>
| PCI: delay configuration of SRIOV capability
| The SRIOV capability, namely page size and total_vfs of a device are
| configured during enumeration phase of the device. This can potentially
| interfere with the PCI operations of the platform, if the IOV capability
| of the device is not enabled.
Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Acked-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
Fixes PCI device detection on IBM xSeries IBM 3850 M2 / x3950 M2
when using ACPI resources (_CRS).
This is default, a manual workaround (without this patch)
would be pci=nocrs boot param.
V2: Add dev_warn if the workaround is hit. This should reveal
how common such setups are (via google) and point to possible
problems if things are still not working as expected.
-> Suggested by Jan Beulich.
Cc: stable@vger.kernel.org
Tested-by: garyhade@us.ibm.com
Signed-off-by: Yinghai Lu <yinghai.lu@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
|
|
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (31 commits)
gma500: Fix suspend/resume functions
drm/exynos: fixed pm feature for fimd module.
MAINTAINERS: added maintainer entry for Exynos DRM Driver.
drm/exynos: fixed build dependency for DRM_EXYNOS_FIMD
drm/exynos: fix build dependency for DRM_EXYNOS_HDMI
drm/exynos: use release_mem_region instead of release_resource
agp: fix scratch page cleanup
drm/i915: fixup forcewake spinlock fallout in drpc debugfs function
drm/i915: debugfs: show semaphore registers also on gen7
drm/i915: allow userspace forcewake references also on gen7
drm/i915: Re-enable gen7 RC6 and GPU turbo after resume.
drm/i915: Correct debugfs printout for RC1e.
Revert "drm/i915: Work around gen7 BLT ring synchronization issues."
drm/i915: rip out the HWSTAM missed irq workaround
drm/i915: paper over missed irq issues with force wake voodoo
drm/i915: Hold gt_lock across forcewake register reads
drm/i915: Hold gt_lock during reset
drm/i915: Move reset forcewake processing to gen6_do_reset
drm/i915: protect force_wake_(get|put) with the gt_lock
drm/i915: convert force_wake_get to func pointer in the gpu reset code
...
|
|
Both the suspend and resume functions incorrectly set psbfb =
to_psb_fb(NULL) outside of the loop over all of the framebuffers. Fix
this by moving the assignment of psbfb inside the loop and removing the
initialisation of fb.
Signed-off-by: Ryan Mallon <rmallon@gmail.com>
Acked-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
this patch separates fimd specific power on/off function from pm function
and the pm interfaces will call that function for power on or off.
and also removes unnecessary codes of resume function.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
|
|
FB based FIMD and DRM based FIMD drivers use same hardware
so with this patch, only one of them would be selected.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
|
|
DRM_EXYNOS_HDMI driver and VIDEO_SAMSUNG_S5P_TV driver should be
not enabled at once because they use same HW blocks. So dependency
for DRM_EXYNOS_HDMI is fixed to check VIDEO_SAMSUNG_S5P_TV=n.
Signed-off-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
|
|
To make a api pair of request_mem_region and release_mem_region,
release_mem_region is used instead of release_resource.
Signed-off-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
[media] cinergyT2-fe: Fix bandwdith settings
[media] V4L: atmel-isi: add clk_prepare()/clk_unprepare() functions
[media] cxd2820r: sleep on DVB-T/T2 delivery system switch
[media] anysee: fix CI init
[media] cxd2820r: remove unused parameter from cxd2820r_attach
[media] cxd2820r: fix dvb_frontend_ops
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:
m68k: Fix assembler constraint to prevent overeager gcc optimisation
mac_esp: rename irq
mac_scsi: dont enable mac_scsi irq before requesting it
macfb: fix black and white modes
m68k/irq: Remove obsolete IRQ_FLG_* definitions
Fix up trivial conflict in arch/m68k/kernel/process_mm.c as per Geert.
|
|
In error cleanup of agp_backend_initialize() and in agp_backend_cleanup(),
agp_destroy_page() is passed virtual address of the scratch page. This
leads to a kernel warning if the initialization fails (or upon regular
cleanup) as pointer to struct page should be passed instead.
Signed-off-by: Michal Kubecek <mkubecek@suse.cz>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
into drm-fixes
* 'drm-intel-fixes' of git://people.freedesktop.org/~keithp/linux: (24 commits)
drm/i915: fixup forcewake spinlock fallout in drpc debugfs function
drm/i915: debugfs: show semaphore registers also on gen7
drm/i915: allow userspace forcewake references also on gen7
drm/i915: Re-enable gen7 RC6 and GPU turbo after resume.
drm/i915: Correct debugfs printout for RC1e.
Revert "drm/i915: Work around gen7 BLT ring synchronization issues."
drm/i915: rip out the HWSTAM missed irq workaround
drm/i915: paper over missed irq issues with force wake voodoo
drm/i915: Hold gt_lock across forcewake register reads
drm/i915: Hold gt_lock during reset
drm/i915: Move reset forcewake processing to gen6_do_reset
drm/i915: protect force_wake_(get|put) with the gt_lock
drm/i915: convert force_wake_get to func pointer in the gpu reset code
drm/i915: sprite init failure on pre-SNB is not a failure
drm/i915: VBT Parser cleanup for eDP block
drm/i915: mask transcoder select bits before setting them on LVDS
drm/i915: Add Clientron E830 to the ignore LVDS list
CHROMIUM: i915: Add DMI override to skip CRT initialization on ZGB
drm/i915: handle 3rd pipe
drm/i915: simplify pipe checking
...
|
|
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
drm/ttm: fix two regressions since move_notify changes
drm/radeon: avoid deadlock if GPU lockup is detected in ib_pool_get
drm/radeon: silence out possible lock dependency warning
drm: Fix authentication kernel crash
gma500: Fix shmem mapping
drm/radeon/kms: refine TMDS dual link checks
drm/radeon/kms: use drm_detect_hdmi_monitor for picking encoder mode
drm/radeon/kms: rework modeset sequence for DCE41 and DCE5
drm/radeon/kms: move panel mode setup into encoder mode set
drm/radeon/kms: move disp eng pll setup to init path
drm/radeon: finish getting bios earlier
drm/radeon: fix invalid memory access in radeon_atrm_get_bios()
drm/radeon/kms: add some missing semaphore init
drm/radeon/kms: Add an MSI quirk for Dell RS690
gpu, drm, sis: Don't return uninitialized variable from sis_driver_load()
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap
A fairly simple bugfix for a WARN_ON() which was triggered in the cache
reset support as a result of some subsequent work. There's only one
mainline user for the code path that's updated right now (wm8994) so
should be low risk.
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap:
regmap: Reset cache status when reinitialsing the cache
|
|
Both changes in dc97b3409a790d2a21aac6e5cdb99558b5944119 cause serious
regressions in the nouveau driver.
move_notify() was originally able to presume that bo->mem is the old node,
and new_mem is the new node. The above commit moves the call to
move_notify() to after move() has been done, which means that now, sometimes,
new_mem isn't the new node at all, bo->mem is, and new_mem points at a
stale, possibly-just-been-killed-by-move node.
This is clearly not a good situation. This patch reverts this change, and
replaces it with a cleanup in the move() failure path instead.
The second issue is that the call to move_notify() from cleanup_memtype_use()
causes the TTM ghost objects to get passed into the driver. This is clearly
bad as the driver knows nothing about these "fake" TTM BOs, and ends up
accessing uninitialised memory.
I worked around this in nouveau's move_notify() hook by ensuring the BO
destructor was nouveau's. I don't particularly like this solution, and
would rather TTM never pass the driver these objects. However, I don't
clearly understand the reason why we're calling move_notify() here anyway
and am happy to work around the problem in nouveau instead of breaking the
behaviour expected by other drivers.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Jerome Glisse <j.glisse@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
My forcewake spinlock patches have a functional conflict with Ben
Widawsky's gen6 drpc support for debugfs. Result was a benign warning
about trying to read an non-atomic variabla with atomic_read.
Note that the entire check is racy anyway and purely informational.
Also update it to reflect the forcewake voodoo changes, the kernel can
now also hold onto a forcewake reference for longer times.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Keith Packard <keithp@keithp.com>
|
|
If GPU lockup is detected in ib_pool get we are holding the ib_pool
mutex that will be needed by the GPU reset code. As ib_pool code is
safe to be reentrant from GPU reset code we should not block if we
are trying to get the ib pool lock on the behalf of the same userspace
caller, thus use the radeon_mutex_lock helper.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
Silence out the lock dependency warning by moving bo allocation out
of ib mutex protected section. Might lead to useless temporary
allocation but it's not harmful as such things only happen at
initialization.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
If the master tries to authenticate a client using drm_authmagic and
that client has already closed its drm file descriptor,
either wilfully or because it was terminated, the
call to drm_authmagic will dereference a stale pointer into kmalloc'ed memory
and corrupt it.
Typically this results in a hard system hang.
This patch fixes that problem by removing any authentication tokens
(struct drm_magic_entry) open for a file descriptor when that file
descriptor is closed.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
