- 07 5月, 2007 9 次提交
-
-
由 Roland Dreier 提交于
The semantics defined by the InfiniBand specification say that completion events are only generated when a completions is added to a completion queue (CQ) after completion notification is requested. In other words, this means that the following race is possible: while (CQ is not empty) ib_poll_cq(CQ); // new completion is added after while loop is exited ib_req_notify_cq(CQ); // no event is generated for the existing completion To close this race, the IB spec recommends doing another poll of the CQ after requesting notification. However, it is not always possible to arrange code this way (for example, we have found that NAPI for IPoIB cannot poll after requesting notification). Also, some hardware (eg Mellanox HCAs) actually will generate an event for completions added before the call to ib_req_notify_cq() -- which is allowed by the spec, since there's no way for any upper-layer consumer to know exactly when a completion was really added -- so the extra poll of the CQ is just a waste. Motivated by this, we add a new flag "IB_CQ_REPORT_MISSED_EVENTS" for ib_req_notify_cq() so that it can return a hint about whether the a completion may have been added before the request for notification. The return value of ib_req_notify_cq() is extended so: < 0 means an error occurred while requesting notification == 0 means notification was requested successfully, and if IB_CQ_REPORT_MISSED_EVENTS was passed in, then no events were missed and it is safe to wait for another event. > 0 is only returned if IB_CQ_REPORT_MISSED_EVENTS was passed in. It means that the consumer must poll the CQ again to make sure it is empty to avoid the race described above. We add a flag to enable this behavior rather than turning it on unconditionally, because checking for missed events may incur significant overhead for some low-level drivers, and consumers that don't care about the results of this test shouldn't be forced to pay for the test. Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Michael S. Tsirkin 提交于
Add a num_comp_vectors member to struct ib_device and extend ib_create_cq() to pass in a comp_vector parameter -- this parallels the userspace libibverbs API. Update all hardware drivers to set num_comp_vectors to 1 and have all ULPs pass 0 for the comp_vector value. Pass the value of num_comp_vectors to userspace rather than hard-coding a value of 1. We want multiple CQ event vector support (via MSI-X or similar for adapters that can generate multiple interrupts), but it's not clear how many vectors we want, or how we want to deal with policy issues such as how to decide which vector to use or how to set up interrupt affinity. This patch is useful for experimenting, since no core changes will be necessary when updating a driver to support multiple vectors, and we know that we want to make at least these changes anyway. Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Ralph Campbell 提交于
Fix a problem where simple ACKs can be sent ahead of RDMA read responses thus implicitly NAKing the RDMA read. Signed-off-by: NRalph Campbell <ralph.cambpell@qlogic.com> Signed-off-by: NRobert Walsh <robert.walsh@qlogic.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Ralph Campbell 提交于
Fix a missing unlock in ipath_rc_rcv_resp() and remove an extra unlock from ipath_rc_rcv_error(). Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Roland Dreier 提交于
Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Roland Dreier 提交于
Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Ishai Rabinovitz 提交于
Add an orig_dgid attribute in sysfs for SRP scsi_hosts, so that userspace can tell what the original dgid value written to the add_target file was, even if the connection is redirected to a different port while connecting. Signed-off-by: NIshai Rabinovitz <ishai@mellanox.co.il> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Michael S. Tsirkin 提交于
The IPoIB CM spec allows the use of a single connection in both active->passive and passive->active directions. The current Linux code uses one connection for both directions, but if another node only uses one connection for both directions, we oops when we try to look up the passive connection. Fix by checking that qp_context is non-NULL before dereferencing it. Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
-
由 Steve Wise 提交于
The HW now posts 2 ABORT_RPL and/or PEER_ABORT_REQ messages. We need to handle them by silenty dropping the 1st but mark that we're ready for the final message. This plugs some close races between the uP and HW. Also update the minimum required firmware version. Signed-off-by: NSteve Wise <swise@opengridcomputing.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 01 5月, 2007 8 次提交
-
-
由 Steve Wise 提交于
Signed-off-by: NSteve Wise <swise@opengridcomputing.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Steve Wise 提交于
Signed-off-by: NSteve Wise <swise@opengridcomputing.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Steve Wise 提交于
Fix TERMINATE layer, type, and ecode values based on conformance testing. Signed-off-by: NSteve Wise <swise@opengridcomputing.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Michael S. Tsirkin 提交于
If skb allocation fails when we start the device, we call ipoib_cm_dev_stop() even though ipoib_cm_dev_open() did not run to completion, so we pass an invalid pointer to ib_destroy_cm_id and get an oops. Fix by clearing cm.id on error, and testing it in cm_dev_stop(). This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=561> Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Robert Walsh 提交于
Fix the pending mmap code so it doesn't corrupt the list of pending mmaps and crash the machine when pending mmaps are destroyed without first being mapped. Also, remove an unused variable, and use standard kernel lists instead of our own homebrewed linked list implementation to keep the pending mmap list. Signed-off-by: NRobert Walsh <robert.walsh@qlogic.com> Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Michael S. Tsirkin 提交于
With mthca, RC QPs can starve each other and even UD QPs on the same hardware schedule queue. As a result, userspace MPI can starve e.g. IPoIB traffic, with netdev watchdog warnings getting printed out, and TCP connections getting stuck or failing. Reduce the chance of this happening by using three separate hardware schedule queues: one for userspace RC QPs, one for kernel RC QPs, and one for all other QPs. Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Ralph Campbell 提交于
This fixes a problem which causes too many RC timeouts and retransmits. Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Ralph Campbell 提交于
This patch fixes the problem reported by Bernd Schubert <bs@q-leap.de> with kernel debug options enabled: BUG: at kernel/lockdep.c:1860 trace_hardirqs_on() This was caused by using spin_lock_irq()/spin_unlock_irq() from interrupt context. Fix all the places that might be called from interrupts to use spin_lock_irqsave()/spin_unlock_irqrestore(). Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 28 4月, 2007 23 次提交
-
-
由 Linus Torvalds 提交于
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: sis900: Allocate rx replacement buffer before rx operation usb-net/pegasus: simplify carrier detection
-
由 Neil Horman 提交于
Just found a hole in my last patch. It was reported to me that shortly after we integrated this patch. The report was of an oops that took place inside of netif_rx when using the sis900 driver. Looking at my origional patch I noted that there was a spot between the new skb_alloc and the refill_rx_ring label where skb got reassigned to the pointer currently held in the rx_ring for the purposes of receiveing the frame. The result of this is however that the buffer that gets passed to netif_rx (if it is called), then gets placed right back into the rx_ring. So if you receive frames fast enough the skb being processed by the network stack can get corrupted. The reporter is testing out the fix I've written for this below (I'm not near my hardware at the moment to test myself), but I wanted to post it for review ASAP. I'll post test results when I hear them, but I think this is a pretty straightforward fix. It just uses a separate pointer to do the rx operation, so that we don't improperly reassign the pointer that we use to refill the rx ring. Signed-off-by: NNeil Horman <nhorman@tuxdriver.com> Signed-off-by: NJeff Garzik <jeff@garzik.org>
-
由 Dan Williams 提交于
Simplify pegasus carrier detection; rely only on the periodic MII polling. Reverts pieces of c43c49bd. Signed-off-by: NDan Williams <dcbw@redhat.com> Signed-off-by: NJeff Garzik <jeff@garzik.org>
-
由 Linus Torvalds 提交于
* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6: [SCSI] esp_scsi.c: Fix compilation.
-
由 Linus Torvalds 提交于
* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: [NET]: Fix networking compilation errors [AF_RXRPC/AFS]: Arch-specific fixes. [AFS]: Fix VLocation record update wakeup [NET]: Revert sk_buff walker cleanups.
-
git://git.infradead.org/mtd-2.6由 Linus Torvalds 提交于
* git://git.infradead.org/mtd-2.6: (46 commits) [MTD] [MAPS] drivers/mtd/maps/ck804xrom.c: convert pci_module_init() [MTD] [NAND] CM-x270 MTD driver [MTD] [NAND] Wrong calculation of page number in nand_block_bad() [MTD] [MAPS] fix plat-ram printk format [JFFS2] Fix compr_rubin.c build after include file elimination. [JFFS2] Handle inodes with only a single metadata node with non-zero isize [JFFS2] Tidy up licensing/copyright boilerplate. [MTD] [OneNAND] Exit loop only when column start with 0 [MTD] [OneNAND] Fix access the past of the real oobfree array [MTD] [OneNAND] Update Samsung OneNAND official URL [JFFS2] Better fix for all-zero node headers [JFFS2] Improve read_inode memory usage, v2. [JFFS2] Improve failure mode if inode checking leaves unchecked space. [JFFS2] Fix cross-endian build. [MTD] Finish conversion mtd_blkdevs to use the kthread API [JFFS2] Obsolete dirent nodes immediately on unlink, where possible. Use menuconfig objects: MTD [MTD] mtd_blkdevs: Convert to use the kthread API [MTD] Fix fwh_lock locking [JFFS2] Speed up mount for directly-mapped NOR flash ...
-
由 Adrian Bunk 提交于
This patch converts the pci_module_init() usage to pci_register_driver(). It's currently #if 0'ed, but still not a bad idea to change it. Signed-off-by: NAdrian Bunk <bunk@stusta.de> Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>
-
由 David Howells 提交于
Fix miscellaneous networking compilation errors. (*) Export ktime_add_ns() for modules. (*) wext_proc_init() should have an ANSI declaration. Signed-off-by: NDavid Howells <dhowells@redhat.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 David Howells 提交于
Fixes for various arch compilation problems: (*) Missing module exports. (*) Variable name collision when rxkad and af_rxrpc both built in (rxrpc_debug). (*) Large constant representation problem (AFS_UUID_TO_UNIX_TIME). (*) Configuration dependencies. (*) printk() format warnings. Signed-off-by: NDavid Howells <dhowells@redhat.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 David Howells 提交于
Fix the wakeup transitions after a VLocation record update completes one way or another. This builds on Dave Miller's partial fix. Also move wakeups outside the spinlocked sections. Signed-off-by: NDavid Howells <dhowells@redhat.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 David S. Miller 提交于
This reverts eefa3906 The simplification made in that change works with the assumption that the 'offset' parameter to these functions is always positive or zero, which is not true. It can be and often is negative in order to access SKB header values in front of skb->data. Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Alexey Dobriyan 提交于
irqreturn.h for irqreturn_t and dma_addr_t being u128 warnings ;-) Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Mike Rapoport 提交于
This patch provides MTD support for NAND flash devices on CM-x270 modules. Signed-off-by: NMike Rapoport <mike@compulab.co.il> Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>
-
由 Linus Torvalds 提交于
* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/usb-2.6: (78 commits) USB: update MAINAINERS and CREDITS for Freescale USB driver USB: update gadget files for fsl_usb2_udc driver USB: add Freescale high-speed USB SOC device controller driver USB: quirk for broken suspend of IT8152F/G USB: iowarrior.c: timeouts too small in usb_control_msg calls USB: dell device id for option.c USB: Remove Huawei unusual_devs entry USB: CP2101 New Device IDs USB: add picdem device to ldusb usbfs micro optimitation USB: remove ancient/broken CRIS hcd usb ethernet gadget, workaround network stack API glitch USB: add "busnum" attribute for USB devices USB: cxacru: ADSL state management usbatm: Detect usb device shutdown and ignore failed urbs USB: Remove duplicate define of OHCI_QUIRK_ZFMICRO USB: BandRich BandLuxe HSDPA Data Card Driver USB gadget rndis: fix struct rndis_packet_msg_type unaligned bug USB Elan FTDI: check for driver registration status USB: sierra: add more checks on shutdown ...
-
ssh://master.kernel.org/pub/scm/linux/kernel/git/mchehab/v4l-dvb由 Linus Torvalds 提交于
* 'master' of ssh://master.kernel.org/pub/scm/linux/kernel/git/mchehab/v4l-dvb: (184 commits) V4L/DVB (5563): Radio-maestro.c Replace radio_ioctl to use video_ioctl2 V4L/DVB (5562): Radio-gemtek-pci.c Replace gemtek_pci_ioctl to use video_ioctl2 V4L/DVB (5560): Ivtv: fix incorrect bitwise-and for command flags. V4L/DVB (5558): Opera: use 7-bit i2c addresses V4L/DVB (5557): Cafe_ccic: check return value of pci_enable_device V4L/DVB (5556): Radio-gemtek.c Replace gemtek_ioctl to use video_ioctl2 V4L/DVB (5555): Radio-aimslab.c Replace rt_ioctl to use video_ioctl2 V4L/DVB (5554): Fix: vidioc_g_parm were not zeroing the memory V4L/DVB (5553): Replace typhoon_do_ioctl to use video_ioctl2 V4L/DVB (5552): Plan-b: Switch to refcounting PCI API V4L/DVB (5551): Plan-b: header change V4L/DVB (5550): Radio-sf16fmi.c Replace fmi_do_ioctl to use video_ioctl2 V4L/DVB (5549): Radio-sf16fmr2.c Replace fmr2_do_ioctl to use video_ioctl2 V4L/DVB (5548): Fix v4l2 buffer to the length V4L/DVB (5547): Add ENUM_FRAMESIZES and ENUM_FRAMEINTERVALS ioctls V4L/DVB (5546): Radio-terratec.c Replace tt_do_ioctl to use video_ioctl2 V4L/DVB (5545): Saa7146: Release capture buffers on device close V4L/DVB (5544): Budget-av: Make inversion setting configurable, add KNC ONE V1.0 card V4L/DVB (5543): Tda10023: Add support for frontend TDA10023 V4L/DVB (5542): Budget-av: Remove polarity switching of the clock for DVB-C ...
-
由 Li Yang 提交于
Add MAINAINERS and CREDITS entry for Freescale Highspeed USB device driver. Signed-off-by: NLi Yang <leoli@freescale.com> Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Li Yang 提交于
Update gadget_chip.c, ether.c for newly added Freescale Highspeed USB device driver. Signed-off-by: NLi Yang <leoli@freescale.com> Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Li Yang 提交于
Freescale high-speed USB SOC can be found on some Freescale processors among different architectures. It supports both host and device functions. This driver adds its device support for Linux USB Gadget layer. It is tested on MPC8349 and MPC8313, but should work on other platforms with minor tweaks. The driver passed USBCV 1.3 compliance tests. Note that this driver doesn't yet include OTG support. Signed-off-by: NLi Yang <leoli@freescale.com> Signed-off-by: NJiang Bo <tanya.jiang@freescale.com> Signed-off-by: NBruce Schmid <duck@freescale.com> Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Raphael Assenat 提交于
Here's a patch which adds my device to the list. This patch enables the broken suspend quirk for the PCI OHCI controller present in the IT8152F/G RISC-to-PCI Companion Chip. Signed-off-by: NRaphael Assenat <raph@8d.com> Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Eberhard Fahle 提交于
The driver uses usb_control_msg() for exchanging data with the device. When the driver lived freeley _outside_ the kernel tree (pre 2.6.21) the timeouts for these calls where set to 5*HZ for reading, 1HZ for writing. (These timeouts seemed to work fine for all users of the driver, at least nobody complained in the last 2 years. The current code (2.6.21-rc5) removed the 'HZ' from the timeouts and left the driver with 5 jiffies for reading and 1 jiffy for writing. My new machine is fast, but not that fast. The patch also removes a useless debug statement, which was left over from testing a broken firmware version From: Eberhard Fahle <e.fahle@wayoda.org> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Hans Engelen 提交于
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Phil Dibowitz 提交于
Per the Rui Santos and the hardware manufacturers, this actually inhibits useful parts of the hardware. The correct way to use this hardware is with the software at http://www.kanoistika.sk/bobovsky/archiv/umts/ and the manufacturers are also planning on including Linux drivers/material in future revisions. CC: Rui Santos <rsantos@grupopie.com> CC: <johann.wilhelm@student.tugraz.at> CC: <zihan@huawei.com> CC: <wanganyu1983@huawei.com> CC: <dingjianjian@huawei.com> Signed-off-by: NPhil Dibowitz <phil@ipom.com> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-
由 Craig Shelley 提交于
Two new device IDs for CP2101 driver. Signed-off-by: NCraig Shelley <craig@microtron.org.uk> Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
-