summaryrefslogtreecommitdiff
path: root/drivers
AgeCommit message (Collapse)AuthorFilesLines
2008-06-12Merge git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6Linus Torvalds4-13/+65
* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6: USB: don't use reset-resume if drivers don't support it USB: isp1760: Assign resource fields before adding hcd isight_firmware: Avoid crash on loading invalid firmware USB: fix build bug in USB_ISIGHTFW
2008-06-12Merge git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-2.6Linus Torvalds1-0/+1
* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-2.6: kobject: Documentation Spelling Patch dev_set_name: fix missing kernel-doc
2008-06-12Merge branch 'release' of ↵Linus Torvalds20-322/+417
git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6 * 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6: (23 commits) ACPICA: fix stray va_end() caused by mis-merge ACPI: Reject below-freezing temperatures as invalid critical temperatures ACPICA: Fix for access to deleted object <regression> ACPICA: Fix to make _SST method optional ACPICA: Fix for Load operator, load table at the namespace root ACPICA: Ignore ACPI table signature for Load() operator ACPICA: Fix to allow zero-length ASL field declarations ACPI: use memory_read_from_buffer() bay: exit if notify handler cannot be installed dock.c remove trailing printk whitespace proper prototype for acpi_processor_tstate_has_changed() ACPI: handle invalid ACPI SLIT table PNPACPI: use _CRS IRQ descriptor length for _SRS pnpacpi: fix shareable IRQ encode/decode pnpacpi: fix IRQ flag decoding MAINTAINERS: update ACPI homepage ACPI 2.6.26-rc2: Add missing newline to DSDT/SSDT warning message ACPI: EC: Use msleep instead of udelay while waiting for event. thinkpad-acpi: fix LED handling on older ThinkPads thinkpad-acpi: fix initialization error paths ...
2008-06-12USB: don't use reset-resume if drivers don't support itLinus Torvalds1-2/+44
This patch tries to identify which devices are able to accept reset-resume handling, by checking that there is at least one interface driver bound and that all of the drivers have a reset_resume method defined. If these conditions don't hold then during resume processing, the device is logicall disconnected. This is only a temporary fix. Later on we will explicitly unbind drivers that can't handle reset-resumes. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Cc: Oliver Neukum <oliver@neukum.org> Cc: Pavel Machek <pavel@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-06-12USB: isp1760: Assign resource fields before adding hcdNate Case1-4/+4
This fixes the bogus "io mem 0x00000000" message printed during driver init due to hcd->rsrc_start being assigned after the call to usb_add_hcd(). Signed-off-by: Nate Case <ncase@xes-inc.com> Acked-by: Sebastian Siewior <bigeasy@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-06-12isight_firmware: Avoid crash on loading invalid firmwareMatthew Garrett1-7/+16
Different tools generate slightly different formats of the isight firmware. Ensure that the firmware buffer is not overrun, while still ensuring that the correct amount of data is written if trailing data is present. Signed-off-by: Matthew Garrett <mjg@redhat.com> Report-by: Justin Mattock <justinmattock@gmail.com> Tested-by: Justin Mattock <justinmattock@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-06-12USB: fix build bug in USB_ISIGHTFWIngo Molnar1-0/+1
USB: fix build bug in USB_ISIGHTFW -tip tree testing found this build bug: drivers/built-in.o: In function `isight_firmware_load': isight_firmware.c:(.text+0x1ade08): undefined reference to `request_firmware' isight_firmware.c:(.text+0x1adf9c): undefined reference to `release_firmware' select FW_LOADER in USB_ISIGHTFW. From: Ingo Molnar <mingo@elte.hu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-06-12dev_set_name: fix missing kernel-docRandy Dunlap1-0/+1
Fix kernel-doc for new dev_set_name() function: Warning(lin2626-rc5//drivers/base/core.c:767): No description found for parameter 'fmt' Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-06-12ACPICA: fix stray va_end() caused by mis-mergeLen Brown1-1/+1
Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI: Reject below-freezing temperatures as invalid critical temperaturesArjan van de Ven1-2/+9
My laptop thinks that it's a good idea to give -73C as the critical CPU temperature.... which isn't the best thing since it causes a shutdown right at bootup. Temperatures below freezing are clearly invalid critical thresholds so just reject these as such. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Acked-by: Zhang Rui <rui.zhang@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPICA: Fix for access to deleted object <regression>Bob Moore1-2/+2
Fixes problem introduced in 20080123, with fix for Unload operator. Parse tree object can be already deleted; must use the opcode within the WalkState. ACPI: kmemcheck: Caught 16-bit read from freed memory http://bugzilla.kernel.org/show_bug.cgi?id=10669 Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPICA: Fix to make _SST method optionalBob Moore1-3/+5
Fixes a problem introduced in 20080514 where the status of execution of _SST is incorrectly returned to the caller. _SST is optional, and if it is AE_NOT_FOUND, the exception should be ignored. http://www.acpica.org/bugzilla/show_bug.cgi?id=716 Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPICA: Fix for Load operator, load table at the namespace rootBob Moore1-2/+8
This reverts a change introduced in version 20071019. The table is now loaded at the namespace root even though this goes against the ACPI specification. This provides compatibility with other ACPI implementations. Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPICA: Ignore ACPI table signature for Load() operatorBob Moore1-18/+7
Only "SSDT" is acceptable to the ACPI spec, but tables are seen with OEMx and null sigs. Therefore, signature validation is worthless. Apparently MS ACPI accepts such signatures, ACPICA must be compatible. http://bugzilla.kernel.org/show_bug.cgi?id=10454 Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPICA: Fix to allow zero-length ASL field declarationsBob Moore1-4/+1
Allows null field list in Field(), BankField(), and IndexField(). 2.6.26-rc1 regression: ACPI fails to load SDT. - Dell M1530 http://bugzilla.kernel.org/show_bug.cgi?id=10606 Signed-off-by: Bob Moore <robert.moore@intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI: use memory_read_from_buffer()Akinobu Mita1-13/+2
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Acked-by: Zhang Rui <rui.zhang@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12bay: exit if notify handler cannot be installedHolger Macht1-6/+10
If acpi_install_notify_handler() for a bay device fails, the bay driver is superfluous. Most likely, another driver (like libata) is already caring about this device anyway. Furthermore, register_hotplug_dock_device(acpi_handle) from the dock driver must not be called twice with the same handler. This would result in an endless loop consuming 100% of CPU. So clean up and exit. Signed-off-by: Holger Macht <hmacht@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12dock.c remove trailing printk whitespaceTim Pepper1-1/+1
Signed-off-by: Tim Pepper <lnxninja@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12proper prototype for acpi_processor_tstate_has_changed()Adrian Bunk1-1/+0
This patch adds a proper prototype for acpi_processor_tstate_has_changed() in include/acpi/processor.h Signed-off-by: Adrian Bunk <bunk@kernel.org> Cc: Len Brown <lenb@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI: handle invalid ACPI SLIT tableFenghua Yu1-4/+27
This is a SLIT sanity checking patch. It moves slit_valid() function to generic ACPI code and does sanity checking for both x86 and ia64. It sets up node_distance with LOCAL_DISTANCE and REMOTE_DISTANCE when hitting invalid SLIT table on ia64. It also cleans up unused variable localities in acpi_parse_slit() on x86. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12PNPACPI: use _CRS IRQ descriptor length for _SRSBjorn Helgaas1-2/+7
When configuring the resources of an ACPI device, we first evaluate _CRS to get a template of resource descriptors, then fill in the specific resource values we want, and finally evaluate _SRS to actually configure the device. Some resources have optional fields, so the size of encoded descriptors varies depending on the specific values. For example, IRQ descriptors can be either two or three bytes long. The third byte contains triggering information and can be omitted if the IRQ is edge-triggered and active high. The BIOS often assumes that IRQ descriptors in the _SRS buffer use the same format as those in the _CRS buffer, so this patch enforces that constraint. The "Start Dependent Function" descriptor also has an optional byte, but we don't currently encode those descriptors, so I didn't do anything for those. I have tested this patch on a Toshiba Portege 4000. Without the patch, parport_pc claims the parallel port only if I use "pnpacpi=off". This patch makes it work with PNPACPI. This is an extension of a patch by Tom Jaeger: http://bugzilla.kernel.org/show_bug.cgi?id=9487#c42 References: http://bugzilla.kernel.org/show_bug.cgi?id=5832 Enabling ACPI Plug and Play in kernels >2.6.9 kills Parallel support http://bugzilla.kernel.org/show_bug.cgi?id=9487 buggy firmware expects four-byte IRQ resource descriptor (was: Serial port disappears after Suspend on Toshiba R25) http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=1d5b285da1893b90507b081664ac27f1a8a3dc5b related ACPICA fix Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12pnpacpi: fix shareable IRQ encode/decodeBjorn Helgaas1-14/+13
When we encode IRQ resources, we should use the "shareable" flag we got from _PRS rather than guessing based on the IRQ trigger mode. This is based on a patch by Tom Jaeger: http://bugzilla.kernel.org/show_bug.cgi?id=9487#c32 Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12pnpacpi: fix IRQ flag decodingBjorn Helgaas1-4/+12
When decoding IRQ trigger mode and polarity, it is not enough to mask by IORESOURCE_BITS because there are now additional bits defined. For example, if IORESOURCE_IRQ_SHAREABLE was set, we failed to set *triggering and *polarity at all. I can't point to a failure that this patch fixes, but bugs in this area have caused problems when resuming after suspend, for example: http://bugzilla.kernel.org/show_bug.cgi?id=6316 http://bugzilla.kernel.org/show_bug.cgi?id=9487 https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.22/+bug/152187 This is based on a patch by Tom Jaeger: http://bugzilla.kernel.org/show_bug.cgi?id=9487#c32 [rene.herman@keyaccess.nl: fix comment] Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI 2.6.26-rc2: Add missing newline to DSDT/SSDT warning messageAlistair John Strachan1-1/+1
As of recently (probably 2.6.26-rc1) I've been getting the following mangling in the kernel log: [4294014.568167] ACPI: DSDT override uses original SSDTs unless "acpi_no_auto_ssdt"<6>CPU0: Intel(R) Pentium(R) Dual CPU E2160 @ 1.80GHz stepping 0d This is due to a missing newline character in the first message. The following patch against 2.6.26-rc2 fixes it. Please apply. Signed-off-by: Alistair John Strachan <alistair@devzero.co.uk> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI: EC: Use msleep instead of udelay while waiting for event.Alexey Starikovskiy1-1/+1
http://bugzilla.kernel.org/show_bug.cgi?id=10724 Signed-off-by: Alexey Starikovskiy <astarikovskiy@suse.de> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12thinkpad-acpi: fix LED handling on older ThinkPadsHenrique de Moraes Holschuh1-27/+28
The less tested codepaths for LED handling, used on ThinkPads 570, 600e/x, 770e, 770x, A21e, A2xm/p, T20-22, X20 and maybe a few others, would write data to kernel memory it had no business touching, for leds number 3 and above. If one is lucky, that illegal write would cause an OOPS, but chances are it would silently corrupt a byte. The problem was introduced in commit af116101, "ACPI: thinkpad-acpi: add sysfs led class support to thinkpad leds (v3.2)". Fix the bug by refactoring the entire code to be far more obvious on what it wants to do. Also do some defensive "constification". Issue reported by Karol Lewandowski <lmctlx@gmail.com> (he's an lucky guy and got an OOPS instead of silent corruption :-) ). Root cause of the OOPS identified by Adrian Bunk <bunk@kernel.org>. Thanks, Adrian! Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br> Tested-by: Karol Lewandowski <lmctlx@gmail.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12thinkpad-acpi: fix initialization error pathsHenrique de Moraes Holschuh1-207/+230
Rework some subdriver init and exit handlers, in order to fix some initialization error paths that were missing, or broken. Hitting those bugs should be extremely rare in the real world, but should that happen, thinkpad-acpi would fail to dealocate some resources and a reboot might well be needed to be able to load the driver again. Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12thinkpad-acpi: SW_RADIO to SW_RFKILL_ALL renameHenrique de Moraes Holschuh1-2/+2
Rename SW_RADIO to SW_RFKILL_ALL in thinkpad-acpi code and docs, following 5adad0133907790c50283bf03271d920d6897043 "Input: rename SW_RADIO to SW_RFKILL_ALL". Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12cpuidle acpi driver: fix oops on AC<->DCVenkatesh Pallipadi2-11/+42
cpuidle and acpi driver interaction bug with the way cpuidle_register_driver() is called. Due to this bug, there will be oops on AC<->DC on some systems, where they support C-states in one DC and not in AC. The current code does ON BOOT: Look at CST and other C-state info to see whether more than C1 is supported. If it is, then acpi processor_idle does a cpuidle_register_driver() call, which internally enables the device. ON CST change notification (AC<->DC) and on suspend-resume: acpi driver temporarily disables device, updates the device with any new C-states, and reenables the device. The problem is is on boot, there are no C2, C3 states supported and we skip the register. Later on AC<->DC, we may get a CST notification and we try to reevaluate CST and enabled the device, without actually registering it. This causes breakage as we try to create /sys fs sub directory, without the parent directory which is created at register time. Thanks to Sanjeev for reporting the problem here. http://bugzilla.kernel.org/show_bug.cgi?id=10394 Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-12ACPI: Disable Fixed_RTC event when installing RTC handlerZhao Yakui2-0/+12
The Fixed_RTC event should be disabled when installing RTC handler. Only when RTC alarm is set will it be enabled again. If it is not disabled, maybe some machines will be powered on automatically after the system is shutdown even when the RTC alarm is not set. http://bugzilla.kernel.org/show_bug.cgi?id=10010 Signed-off-by: Zhao Yakui <yakui.zhao@intel.com> Signed-off-by: Zhang Rui <rui.zhang@intel.com> Signed-off-by: Len Brown <len.brown@intel.com>
2008-06-11Merge branch 'kvm-updates-2.6.26' of ↵Linus Torvalds1-0/+1
git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm * 'kvm-updates-2.6.26' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm: KVM: MMU: Fix is_empty_shadow_page() check KVM: MMU: Fix printk() format string KVM: IOAPIC: only set remote_irr if interrupt was injected KVM: MMU: reschedule during shadow teardown KVM: VMX: Clear CR4.VMXE in hardware_disable KVM: migrate PIT timer KVM: ppc: Report bad GFNs KVM: ppc: Use a read lock around MMU operations, and release it on error KVM: ppc: Remove unmatched kunmap() call KVM: ppc: add lwzx/stwz emulation KVM: ppc: Remove duplicate function KVM: s390: Fix race condition in kvm_s390_handle_wait KVM: s390: Send program check on access error KVM: s390: fix interrupt delivery KVM: s390: handle machine checks when guest is running KVM: s390: fix locking order problem in enable_sie KVM: s390: use yield instead of schedule to implement diag 0x44 KVM: x86 emulator: fix hypercall return value on AMD KVM: ia64: fix zero extending for mmio ld1/2/4 emulation in KVM
2008-06-11Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6Linus Torvalds28-339/+311
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (42 commits) net: Fix routing tables with id > 255 for legacy software sky2: Hold RTNL while calling dev_close() s2io iomem annotations atl1: fix suspend regression qeth: start dev queue after tx drop error qeth: Prepare-function to call s390dbf was wrong qeth: reduce number of kernel messages qeth: Use ccw_device_get_id(). qeth: layer 3 Oops in ip event handler virtio: use callback on empty in virtio_net virtio: virtio_net free transmit skbs in a timer virtio: Fix typo in virtio_net_hdr comments virtio_net: Fix skb->csum_start computation ehea: set mac address fix sfc: Recover from RX queue flush failure add missing lance_* exports ixgbe: fix typo forcedeth: msi interrupts ipsec: pfkey should ignore events when no listeners pppoe: Unshare skb before anything else ...
2008-06-11smc91x: fix build error from the SMC_GET_MAC_ADDR API changeBryan Wu1-4/+4
Cc: Jeff Garzik <jeff@garzik.org> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Bryan Wu <cooloney@kernel.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-06-11Merge branch 'davem-fixes' of ↵David S. Miller15-188/+150
master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
2008-06-11sky2: Hold RTNL while calling dev_close()Ben Hutchings1-0/+2
dev_close() must be called holding the RTNL. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11s2io iomem annotationsAl Viro1-4/+6
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11atl1: fix suspend regressionJay Cliburn1-17/+0
Using vendor magic to force the PHY into power save mode breaks suspend. It isn't needed anyway, so remove it. Tested-by: Avuton Olrich <avuton@gmail.com> Signed-off-by: Jay Cliburn <jacliburn@bellsouth.net> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11qeth: start dev queue after tx drop errorFrank Blaschka2-0/+2
In case the xmit function drop out with an error, we have to wake the netdevice queue to start another xmit. Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11qeth: Prepare-function to call s390dbf was wrongPeter Tiedemann1-3/+5
Prepare-function to call s390dbf was wrong handling variable arguments. This worked as macro but not as function any more. Now using va_list processing. Signed-off-by: Peter Tiedemann <ptiedem@de.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11qeth: reduce number of kernel messagesFrank Blaschka6-131/+42
Remove unnecessary messages. Write important debug information to s390dbf. Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11qeth: Use ccw_device_get_id().Cornelia Huck2-15/+7
Get the devno from the ccw device via ccw_device_get_id() instead of parsing the bus_id. Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11qeth: layer 3 Oops in ip event handlerFrank Blaschka1-1/+1
The ip event handler may present us non qeth network interfaces. Add qeth card pointer check. Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11virtio: use callback on empty in virtio_netRusty Russell1-5/+17
virtio_net uses a timer to free old transmitted packets, rather than leaving callbacks enabled all the time. If the host promises to always notify us when the transmit ring is empty, we can free packets at that point and avoid the timer. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11virtio: virtio_net free transmit skbs in a timerMark McLoughlin1-2/+26
virtio_net currently only frees old transmit skbs just before queueing new ones. If the queue is full, it then enables interrupts and waits for notification that more work has been performed. However, a side-effect of this scheme is that there are always xmit skbs left dangling when no new packets are sent, against the Documentation/networking/driver.txt guideline: "... it is not allowed for your TX mitigation scheme to let TX packets "hang out" in the TX ring unreclaimed forever if no new TX packets are sent." Add a timer to ensure that any time we queue new TX skbs, we will shortly free them again. This fixes an easily reproduced hang at shutdown where iptables attempts to unload nf_conntrack and nf_conntrack waits for an skb it is tracking to be freed, but virtio_net never frees it. Signed-off-by: Mark McLoughlin <markmc@redhat.com> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11virtio_net: Fix skb->csum_start computationMark McLoughlin1-3/+5
hdr->csum_start is the offset from the start of the ethernet header to the transport layer checksum field. skb->csum_start is the offset from skb->head. skb_partial_csum_set() assumes that skb->data points to the ethernet header - i.e. it computes skb->csum_start by adding the headroom to hdr->csum_start. Since eth_type_trans() skb_pull()s the ethernet header, skb_partial_csum_set() should be called before eth_type_trans(). (Without this patch, GSO packets from a guest to the world outside the host are corrupted). Signed-off-by: Mark McLoughlin <markmc@redhat.com> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11ehea: set mac address fixJan-Bernd Themann1-6/+10
eHEA has to call firmware functions in order to change the mac address of a logical port. This patch checks if the logical port is up when calling the register / deregister mac address calls. If the port is down these firmware calls would fail and are therefore not executed. Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11sfc: Recover from RX queue flush failureSteve Hodgson1-1/+3
RX queue flush can fail if traffic continues to arrive. Recover by performing an invisible reset. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11add missing lance_* exportsAdrian Bunk1-1/+5
This patch fixes the following build error: <-- snip --> ... Building modules, stage 2. MODPOST 1203 modules ERROR: "lance_open" [drivers/net/mvme147.ko] undefined! ERROR: "lance_close" [drivers/net/mvme147.ko] undefined! ERROR: "lance_tx_timeout" [drivers/net/mvme147.ko] undefined! ERROR: "lance_set_multicast" [drivers/net/mvme147.ko] undefined! ERROR: "lance_start_xmit" [drivers/net/mvme147.ko] undefined! ... make[2]: *** [__modpost] Error 1 <-- snip --> Reported-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11ixgbe: fix typoJeff Kirsher1-2/+2
Define names were accidently transposed. Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-06-11forcedeth: msi interruptsAyaz Abdulla1-0/+20
Add a workaround for lost MSI interrupts. There is a race condition in the HW in which future interrupts could be missed. The workaround is to toggle the MSI irq mask. Added cleanup based on comments from Andrew Morton. Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com> Cc: Manfred Spraul <manfred@colorfullife.com> Cc: Jeff Garzik <jeff@garzik.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
generated by cgit v1.2.3 (git 2.39.0) at 2024-12-30 07:39:20 +0300