- 28 Jun, 2005 24 commits
-
-
David S. Miller authored
After seeing, at best, "guesses" as to the following kind of information in several drivers, I decided that we really need a way for platforms to specifically give advice in this area for what works best with their PCI controller implementation. Basically, this new interface gives DMA bursting advice on PCI. There are three forms of the advice: 1) Burst as much as possible, it is not necessary to end bursts on some particular boundary for best performance. 2) Burst on some byte count multiple. A DMA burst to some multiple of number of bytes may be done, but it is important to end the burst on an exact multiple for best performance. The best example of this I am aware of are the PPC64 PCI controllers, where if you end a burst mid-cacheline then chip has to refetch the data and the IOMMU translations which hurts performance a lot. 3) Burst on a single byte count multiple. Bursts shall end exactly on the next multiple boundary for best performance. Sparc64 and Alpha's PCI controllers operate this way. They disconnect any device which tries to burst across a cacheline boundary. Actually, newer sparc64 PCI controllers do not have this behavior. That is why the "pdev" is passed into the interface, so I can add code later to check which PCI controller the system is using and give advice accordingly. Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
Greg Kroah-Hartman <gregkh@suse.de>
-
Michael Ellerman authored
This is an updated version of Ben's fix-pci-mmap-on-ppc-and-ppc64.patch which is in 2.6.12-rc4-mm1. It fixes the patch to work on PPC iSeries, removes some debug printks at Ben's request, and incorporates your fix-pci-mmap-on-ppc-and-ppc64-fix.patch also. Originally from Benjamin Herrenschmidt <benh@kernel.crashing.org> This patch was discussed at length on linux-pci and so far, the last iteration of it didn't raise any comment. It's effect is a nop on architecture that don't define the new pci_resource_to_user() callback anyway. It allows architecture like ppc who put weird things inside of PCI resource structures to convert to some different value for user visible ones. It also fixes mmap'ing of IO space on those archs. Signed-off-by:
Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by:
Michael Ellerman <michael@ellerman.id.au> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
-
Kenji Kaneshige authored
This patch adds PCI based I/O xAPIC hot-add support to ACPIPHP driver. When PCI root bridge is hot-added, all PCI based I/O xAPICs under the root bridge are hot-added by this patch. Hot-remove support is TBD. Signed-off-by:
Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com> Signed-off-by:
-
Kenji Kaneshige authored
This is an ia64 implementation of acpi_register_ioapic() and acpi_unregister_ioapic() interfaces. Signed-off-by:
-
Kenji Kaneshige authored
This patch adds the following new interfaces for I/O xAPIC hotplug. The implementation of these interfaces depends on each architecture. o int acpi_register_ioapic(acpi_handle handle, u64 phys_addr, u32 gsi_base); This new interface is to add a new I/O xAPIC specified by phys_addr and gsi_base pair. phys_addr is the physical address to which the I/O xAPIC is mapped and gsi_base is global system interrupt base of the I/O xAPIC. acpi_register_ioapic returns 0 on success, or negative value on error. o int acpi_unregister_ioapic(acpi_handle handle, u32 gsi_base); This new interface is to remove a I/O xAPIC specified by gsi_base. acpi_unregister_ioapic returns 0 on success, or negative value on error. Signed-off-by: -
Rajesh Shah authored
Current acpiphp code does not distinguish between the physical presence and power state of a device/slot. That is, if a device has to be disabled, it also tries to physically ejects the device. This patch decouples power state from physical presence. You can now echo to the corresponding sysfs power control file to repeatedly enable and disable a device without having to physically re-insert it. Signed-off-by:
Rajesh Shah <rajesh.shah@intel.com> Signed-off-by:
-
Rajesh Shah authored
acpiphp changes to support acpi based root bridge hot-add. Signed-off-by:
-
Rajesh Shah authored
Earlier I reported that Matthew's acpiphp rewrite had problem in powering down slot on my i386 system. The following patch is needed to get the acpiphp rewrite properly powering down the slot. Signed-off-by:
Dely Sy <dely.l.sy@intel.com> Signed-off-by:
-
Rajesh Shah authored
A root bridge may not have directly attached hotpluggable slots under it. Instead, it may have p2p bridges with slots under it. In this case, we need to clean up the p2p bridges and slots properly too. Patch below applies on top of the original patch, and fixes this problem. Without this, acpiphp leaves behind notify handlers on module unload, and subsequent module load attempts don't work properly too. Patch was tested on an ia64 Tiger4 box. Signed-off-by:
-
Rajesh Shah authored
This patch converts acpiphp to use the generic PCI resource assignment code. It's quite large, but most of it is deleting the acpiphp_pci and acpiphp_res files. It's tested on an hp Integrity rx8620 (which won't work without this patch). Testers with other hardware welcomed. Signed-off-by:
Matthew Wilcox <matthew@wil.cx> Signed-off-by:
-
Rajesh Shah authored
Export an acpi interface to get PCI domain/bus/devfn information from the corresponding namespace handle. Used by acpiphp code to transpate the device handle of the hot-plugged root bridge to the corresponding pci location information. Signed-off-by:
-
Rajesh Shah authored
Create new interfaces to recursively add an acpi namespace object to the acpi device list, and recursively start the namespace object. This is needed for ACPI based hotplug of a root bridge hierarchy where the add operation must be performed first and the start operation must be performed separately after the hot-plugged devices have been properly configured. Signed-off-by:
-
Rajesh Shah authored
Read bridge io/mem/pfmem ranges when fixing up the bus so that bus resources are tracked. This is required to properly support pci end device and bridge hotplug. Signed-off-by:
-
Rajesh Shah authored
When hot-plugging an I/O hierarchy that contains many bridges and leaf devices, it's possible that there are not enough resources to start all the device present. If we fail to assign a resource, clear the corresponding value in the pci_dev structure, so other code can take corrective action. Signed-off-by:
-
Rajesh Shah authored
When a root bridge hierarchy is hot-plugged, resource requirements for the new devices may be greater than what the root bridge is decoding. In this case, we want to remove devices that did not get needed resources. These devices have been scanned into bus specific lists but not yet added to the global device list. Make sure the pci remove functions can handle this case. Signed-off-by:
-
Rajesh Shah authored
When a pci child bus is created, add it to the parent's children list immediately rather than waiting till pci_bus_add_devices(). For hot-plug bridges/devices, pci_bus_add_devices() may be called much later, after they have been properly configured. In the meantime, this allows us to use the normal pci bus search functions for the hot-plug bridges/buses. Signed-off-by:
-
Rajesh Shah authored
With root bridge and pci bridge hot-plug, new buses and devices can be added or removed at run time. Protect the pci bus and device lists with the pci lock when doing so. Signed-off-by:
-
Rajesh Shah authored
When hot-plugging a root bridge, as we try to assign bus numbers we may find that the hotplugged hieratchy has more PCI to PCI bridges (i.e. bus requirements) than available. Make sure we don't step over an existing bus when that happens. Signed-off-by:
-
Rajesh Shah authored
PCI scan code calls the arch specific pcibios_fixup_bus() each time it scans a new bridge. For root bridge hot-plug, the bridge and it's attached devices may not have been configured properly yet, so it's not safe to claim those resources at this time. This code goes away when we clean up the way pci resources are claimed (in pci_enable_device()), so this is only a stopgap fix. Signed-off-by:
-
Rajesh Shah authored
When checking if a PCI to PCI bridge should be enabled to decode memory and/or IO resources, we need to look at all device resources not just the first 6. This is needed to allow PCI bridges to pass down memory and IO accesses to child devices even when the bridge itself does not consume resources in its PCI BARs. Signed-off-by:
-
Rajesh Shah authored
When you hot-plug a (root) bridge hierarchy, it may have p2p bridges and devices attached to it that have not been configured by firmware. In this case, we need to configure the devices before starting them. This patch separates device start from device scan so that we can introduce the configuration step in the middle. I kept the existing semantics for pci_scan_bus() since there are a huge number of callers to that function. Also, I have no way of testing the changes I made to the parisc files, so this needs review by those folks. Sorry for the massive cross-post, this touches files in many different places. Signed-off-by:
-
Amit Gud authored
Replace pci_find_device() with more safer pci_get_device(). Signed-off-by:
Amit Gud <gud@eth.net> Signed-off-by:
-
Amit Gud authored
Replace pci_find_device() with more safer pci_get_device(). Signed-off-by:
-
jayalk@intworks.biz authored
I updated this to remove unnecessary variable initialization, make check_routing be inline only and not __init, switch to strtoul, and formatting fixes as per Randy Dunlap's recommendations. I updated this to change pirq_table_addr to a long, and to add a warning msg if the PIRQ table wasn't found at the specified address, as per thread with Matthew Wilcox. In our hardware situation, the BIOS is unable to store or generate it's PIRQ table in the F0000h-100000h standard range. This patch adds a pci kernel parameter, pirqaddr to allow the bootloader (or BIOS based loader) to inform the kernel where the PIRQ table got stored. A beneficial side-effect is that, if one's BIOS uses a static address each time for it's PIRQ table, then pirqaddr can be used to avoid the $pirq search through that address block each time at boot for normal PIRQ BIOSes. Signed-off-by:
Jaya Kumar <jayalk@intworks.biz> Signed-off-by:
-
- 27 Jun, 2005 2 commits
-
-
-
Linus Torvalds authored
-
- 26 Jun, 2005 14 commits
-
-
pageexec authored
From: <pageexec@freemail.hu> $subject was fixed in 2.4 already, 2.6 needs it as well. The impact of the bugs is a kernel stack overflow and privilege escalation from CAP_NET_ADMIN via the IP_VS_SO_SET_STARTDAEMON/IP_VS_SO_GET_DAEMON ioctls. People running with 'root=all caps' (i.e., most users) are not really affected (there's nothing to escalate), but SELinux and similar users should take it seriously if they grant CAP_NET_ADMIN to other users. Signed-off-by:
-
Sascha Hauer authored
This patch fixes two bugs in the dm9000 network driver: - Don't read one byte too much in 8bit mode. - release correct resource Signed-off-by:
Jochen Karrer <j.karrer@lightmaze.com> Signed-off-by:
Sascha Hauer <s.hauer@pengutronix.de> Cc: Jeff Garzik <jgarzik@pobox.com> Signed-off-by:
-
Ismail Donmez authored
Patch indents dmfe.txt to look like other docs. It adds a tip about CNET cards using Davicom chipsets. Also it removes parts where it refers to how to build driver out-of-kernel which seems to be cruft from times where the driver was out of the kernel. Signed-off-by:
Ismail Donmez <ismail@kde.org.tr> Signed-off-by:
Domen Puncer <domen@coderock.org> Cc: Jeff Garzik <jgarzik@pobox.com> Signed-off-by:
-
Denis Vlasenko authored
Signed-off-by: -
Pavel Machek authored
Fix int vs. pm_message_t confusion in airo. Should change no code. Signed-off-by:
Pavel Machek <pavel@suse.cz> Cc: Jeff Garzik <jgarzik@pobox.com> Signed-off-by:
-
Maciej W. Rozycki authored
This is a fix for the interrupt handler in the defxx driver to use irqreturn_t. Beside the obvious fix of returning a proper status at all, it actually checks board registers as appropriate for determining if an interrupt has been recorded in the bus-specific interface logic. The patch also includes an obvious one-line fix for SET_NETDEV_DEV needed for the EISA variation, for which I've decided there is no point in sending separately. Signed-off-by:
Maciej W. Rozycki <macro@linux-mips.org> Cc: Jeff Garzik <jgarzik@pobox.com> Signed-off-by:
-
Andrew Morton authored
drivers/net/tulip/dmfe.c: In function `dmfe_parse_srom': drivers/net/tulip/dmfe.c:1805: warning: passing arg 1 of `__le16_to_cpup' from incompatible pointer type drivers/net/tulip/dmfe.c:1817: warning: passing arg 1 of `__le32_to_cpup' from incompatible pointer type drivers/net/tulip/dmfe.c:1817: warning: passing arg 1 of `__le32_to_cpup' from incompatible pointer type This is basically a guess: Cc: Jeff Garzik <jgarzik@pobox.com> Signed-off-by: -
Adrian Bunk authored
The 8129/8130 support is a sub-option that is not visible if the user hasn't enabled the 8139 support. Let's make it a bit easier for users to find the driver for their nic. Signed-off-by:Adrian Bunk <bunk@stusta.de>
-
David S. Miller authored
1) netlink_release() should only decrement the hash entry count if the socket was actually hashed. This was causing hash->entries to underflow, which resulting in all kinds of troubles. On 64-bit systems, this would cause the following conditional to erroneously trigger: err = -ENOMEM; if (BITS_PER_LONG > 32 && unlikely(hash->entries >= UINT_MAX)) goto err; 2) netlink_autobind() needs to propagate the error return from netlink_insert(). Otherwise, callers will not see the error as they should and thus try to operate on a socket with a zero pid, which is very bad. However, it should not propagate -EBUSY. If two threads race to autobind the socket, that is fine. This is consistent with the autobind behavior in other protocols. So bug #1 above, combined with this one, resulted in hangs on netlink_sendmsg() calls to the rtnetlink socket. We'd try to do the user sendmsg() with the socket's pid set to zero, later we do a socket lookup using that pid (via the value we stashed away in NETLINK_CB(skb).pid), but that won't give us the user socket, it will give us the rtnetlink socket. So when we try to wake up the receive queue, we dive back into rtnetlink_rcv() which tries to recursively take the rtnetlink semaphore. Thanks to Jakub Jelink for providing backtraces. Also, thanks to Herbert Xu for supplying debugging patches to help track this down, and also finding a mistake in an earlier version of this fix. Signed-off-by: -
Nishanth Aravamudan authored
Use msleep_interruptible() instead of schedule_timeout() to guarantee the task delays as expected.
-
Adrian Bunk authored
This patch removes some obviously dead code found by the Coverity checker. Signed-off-by: -
Adrian Bunk authored
This patch fixes the LITTLE_ENDIAN #define and a function prototype. Signed-off-by: -
Adrian Bunk authored
This patch contains the follwing cleanups: - make needlessly global code static - remove obsolete Emacs settings Signed-off-by: -
David S. Miller authored
We're using __be16 in userland visible types, so we have to include asm/byteorder.h so that works. Signed-off-by:
-
