1102 #1
Merged
1102 #1
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
On receiving an ACK that covers the loss probe sequence, TLP immediately sets the congestion state to Open, even though some packets are not recovered and retransmisssion are on the way. The later ACks may trigger a WARN_ON check in step D of tcp_fastretrans_alert(), e.g., https://bugzilla.redhat.com/show_bug.cgi?id=989251 The fix is to follow the similar procedure in recovery by calling tcp_try_keep_open(). The sender switches to Open state if no packets are retransmissted. Otherwise it goes to Disorder and let subsequent ACKs move the state to Recovery or Open. Reported-By: Michael Sterrett <[email protected]> Tested-By: Dormando <[email protected]> Signed-off-by: Yuchung Cheng <[email protected]> Acked-by: Neal Cardwell <[email protected]> Signed-off-by: David S. Miller <[email protected]>
RPS support is kind of broken on bnx2x, because only non LRO packets get proper rx queue information. This triggers reorders, as it seems bnx2x like to generate a non LRO packet for segment including TCP PUSH flag : (this might be pure coincidence, but all the reorders I've seen involve segments with a PUSH) 11:13:34.335847 IP A > B: . 415808:447136(31328) ack 1 win 457 <nop,nop,timestamp 3789336 3985797> 11:13:34.335992 IP A > B: . 447136:448560(1424) ack 1 win 457 <nop,nop,timestamp 3789336 3985797> 11:13:34.336391 IP A > B: . 448560:479888(31328) ack 1 win 457 <nop,nop,timestamp 3789337 3985797> 11:13:34.336425 IP A > B: P 511216:512640(1424) ack 1 win 457 <nop,nop,timestamp 3789337 3985798> 11:13:34.336423 IP A > B: . 479888:511216(31328) ack 1 win 457 <nop,nop,timestamp 3789337 3985798> 11:13:34.336924 IP A > B: . 512640:543968(31328) ack 1 win 457 <nop,nop,timestamp 3789337 3985798> 11:13:34.336963 IP A > B: . 543968:575296(31328) ack 1 win 457 <nop,nop,timestamp 3789337 3985798> We must call skb_record_rx_queue() to properly give to RPS (and more generally for TX queue selection on forward path) the receive queue information. Similar fix is needed for skb_mark_napi_id(), but will be handled in a separate patch to ease stable backports. Signed-off-by: Eric Dumazet <[email protected]> Cc: Willem de Bruijn <[email protected]> Cc: Eilon Greenstein <[email protected]> Acked-by: Dmitry Kravkov <[email protected]> Signed-off-by: David S. Miller <[email protected]>
If interrupts happen before napi_enable was called, the driver will not work as expected. Network transmissions are impossible in this state. This bug can be reproduced easily by restarting the network interface in a loop. After some time any network transmissions on the network interface will fail. This patch fixes the bug by enabling napi before enabling the network interface interrupts. Signed-off-by: Markus Pargmann <[email protected]> Acked-by: Peter Korsgaard <[email protected]> Acked-by: Mugunthan V N <[email protected]> Signed-off-by: David S. Miller <[email protected]>
We cap bitrate at YAM_MAXBITRATE in yam_ioctl(), but it could also be negative. I don't know the impact of using a negative bitrate but let's prevent it. Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: David S. Miller <[email protected]>
We overwrite the ->bitrate with the user supplied information on the next line. Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: David S. Miller <[email protected]>
We're trying to re-configure the affinity unconditionally in cpu hotplug callback. This may lead the issue during resuming from s3/s4 since - virt queues haven't been allocated at that time. - it's unnecessary since thaw method will re-configure the affinity. Fix this issue by checking the config_enable and do nothing is we're not ready. The bug were introduced by commit 8de4b2f (virtio-net: reset virtqueue affinity when doing cpu hotplug). Cc: Rusty Russell <[email protected]> Cc: Michael S. Tsirkin <[email protected]> Cc: Wanlong Gao <[email protected]> Acked-by: Michael S. Tsirkin <[email protected]> Reviewed-by: Wanlong Gao <[email protected]> Signed-off-by: Jason Wang <[email protected]> Signed-off-by: David S. Miller <[email protected]>
We used to schedule the refill work unconditionally after changing the number of queues. This may lead an issue if the device is not up. Since we only try to cancel the work in ndo_stop(), this may cause the refill work still work after removing the device. Fix this by only schedule the work when device is up. The bug were introduce by commit 9b9cd80. (virtio-net: fix the race between channels setting and refill) Cc: Rusty Russell <[email protected]> Cc: Michael S. Tsirkin <[email protected]> Signed-off-by: Jason Wang <[email protected]> Signed-off-by: David S. Miller <[email protected]>
This is a QMI device, manufactured by TCT Mobile Phones. A companion patch blacklisting this device's QMI interface in the option.c driver has been sent. Signed-off-by: Enrico Mioso <[email protected]> Signed-off-by: Antonella Pellizzari <[email protected]> Tested-by: Dan Williams <[email protected]> Acked-by: Greg Kroah-Hartman <[email protected]> Signed-off-by: David S. Miller <[email protected]>
…/git/linville/wireless John W. Linville says: ==================== Please pull this batch of fixes intended for the 3.12 stream! For the mac80211 bits, Johannes says: "Jouni fixes a remain-on-channel vs. scan bug, and Felix fixes client TX probing on VLANs." And also: "This time I have two fixes from Emmanuel for RF-kill issues, and fixed two issues reported by Evan Huus and Thomas Lindroth respectively." On top of those... Avinash Patil adds a couple of mwifiex fixes to properly inform cfg80211 about some different types of disconnects, avoiding WARNINGs. Mark Cave-Ayland corrects a pointer arithmetic problem in rtlwifi, avoiding incorrect automatic gain calculations. Solomon Peachy sends a cw1200 fix for locking around calls to cw1200_irq_handler, addressing "lost interrupt" problems. Please let me know if there are problems! ==================== Signed-off-by: David S. Miller <[email protected]>
TCP stack should make sure it owns skbs before mangling them. We had various crashes using bnx2x, and it turned out gso_size was cleared right before bnx2x driver was populating TC descriptor of the _previous_ packet send. TCP stack can sometime retransmit packets that are still in Qdisc. Of course we could make bnx2x driver more robust (using ACCESS_ONCE(shinfo->gso_size) for example), but the bug is TCP stack. We have identified two points where skb_unclone() was needed. This patch adds a WARN_ON_ONCE() to warn us if we missed another fix of this kind. Kudos to Neal for finding the root cause of this bug. Its visible using small MSS. Signed-off-by: Eric Dumazet <[email protected]> Signed-off-by: Neal Cardwell <[email protected]> Cc: Yuchung Cheng <[email protected]> Signed-off-by: David S. Miller <[email protected]>
sk_can_gso() should only be used as a hint in tcp_sendmsg() to build GSO packets in the first place. (As a performance hint) Once we have GSO packets in write queue, we can not decide they are no longer GSO only because flow now uses a route which doesn't handle TSO/GSO. Core networking stack handles the case very well for us, all we need is keeping track of packet counts in MSS terms, regardless of segmentation done later (in GSO or hardware) Right now, if tcp_fragment() splits a GSO packet in two parts, @left and @right, and route changed through a non GSO device, both @left and @right have pcount set to 1, which is wrong, and leads to incorrect packet_count tracking. This problem was added in commit d5ac99a ("[TCP]: skb pcount with MTU discovery") Signed-off-by: Eric Dumazet <[email protected]> Signed-off-by: Neal Cardwell <[email protected]> Signed-off-by: Yuchung Cheng <[email protected]> Reported-by: Maciej Żenczykowski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Commit be4f154 bridge: Clamp forward_delay when enabling STP had a typo when attempting to clamp maximum forward delay. It is possible to set bridge_forward_delay to be higher then permitted maximum when STP is off. When turning STP on, the higher then allowed delay has to be clamed down to max value. CC: Herbert Xu <[email protected]> CC: Stephen Hemminger <[email protected]> Signed-off-by: Vlad Yasevich <[email protected]> Reviewed-by: Veaceslav Falico <[email protected]> Acked-by: Herbert Xu <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Since commit 055560b (serial: at91: distinguish usart and uart) the older products which do not have a name field in their register map are unable to use their serial output. As the main console output is usually the serial interface (aka DBGU) it is pretty unfortunate. So, instead of failing during probe() we just silently configure the serial peripheral as an uart. It allows us to use these serial outputs. The proper solution is proposed in another patch. Signed-off-by: Nicolas Ferre <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
Pull CIFS fixes from Steve French: "Five small cifs fixes (includes fixes for: unmount hang, 2 security related, symlink, large file writes)" * 'for-linus' of git://git.samba.org/sfrench/cifs-2.6: cifs: ntstatus_to_dos_map[] is not terminated cifs: Allow LANMAN auth method for servers supporting unencapsulated authentication methods cifs: Fix inability to write files >2GB to SMB2/3 shares cifs: Avoid umount hangs with smb2 when server is unresponsive do not treat non-symlink reparse points as valid symlinks
…ructs Apply the protections from commit 1b2f148 Author: Dave Airlie <[email protected]> Date: Sat Aug 14 20:20:34 2010 +1000 drm: block userspace under allocating buffer and having drivers overwrite it (v2) to the core ioctl structs as well, for we found one instance where there is a 32-/64-bit size mismatch and were guilty of writing beyond the end of the user's buffer. Signed-off-by: Chris Wilson <[email protected]> Cc: Dave Airlie <[email protected]> Reviewed-by: Ville Syrjälä <[email protected]> Cc: [email protected] Cc: [email protected] Signed-off-by: Dave Airlie <[email protected]>
Pavel Roskin reported that DRM_IOCTL_MODE_GETCONNECTOR was overwritting the 4 bytes beyond the end of its structure with a 32-bit userspace running on a 64-bit kernel. This is due to the padding gcc inserts as the drm_mode_get_connector struct includes a u64 and its size is not a natural multiple of u64s. 64-bit kernel: sizeof(drm_mode_get_connector)=80, alignof=8 sizeof(drm_mode_get_encoder)=20, alignof=4 sizeof(drm_mode_modeinfo)=68, alignof=4 32-bit userspace: sizeof(drm_mode_get_connector)=76, alignof=4 sizeof(drm_mode_get_encoder)=20, alignof=4 sizeof(drm_mode_modeinfo)=68, alignof=4 Fortuituously we can insert explicit padding to the tail of our structures without breaking ABI. Reported-by: Pavel Roskin <[email protected]> Signed-off-by: Chris Wilson <[email protected]> Cc: Dave Airlie <[email protected]> Cc: [email protected] Cc: [email protected] Signed-off-by: Dave Airlie <[email protected]>
* pm-fixes: cpufreq: s3c64xx: Rename index to driver_data intel_pstate: Fix type mismatch warning cpufreq / intel_pstate: Fix max_perf_pct on resume
* acpi-fixes: ACPI / PM: Drop two functions that are not used any more ATA / ACPI: remove power dependent device handling ACPI / power: Drop automaitc resume of power resource dependent devices ACPI: remove /proc/acpi/event from ACPI_BUTTON help ACPI / power: Release resource_lock after acpi_power_get_state() return error
Revert some changes done in 7746383. Revert all changes done in hidinput_calc_abs_res as it mistakingly used "Unit" item exponent nibbles to affect resolution value. This wasn't breaking resolution calculation of relevant axes of any existing devices, though, as they have only one dimension to their units and thus 1 in the corresponding nible. Revert to reading "Unit Exponent" item value as a signed integer in hid_parser_global to fix reading specification-complying values. This fixes resolution calculation of devices complying to the HID standard, including Huion, KYE, Waltop and UC-Logic graphics tablets which have their report descriptors fixed by the drivers. Explanations follow. There are two "unit exponents" in HID specification and it is important not to mix them. One is the global "Unit Exponent" item and another is nibble values in the global "Unit" item. See 6.2.2.7 Global Items. The "Unit Exponent" value is just a signed integer and is used to scale the integer resolution unit values, so fractions can be expressed. The nibbles of "Unit" value are used to select the unit system (nibble 0), and presence of a particular basic unit type in the unit formula and its *exponent* (or power, nibbles 1-6). And yes, the latter is in two complement and zero means absence of the unit type. Taking the representation example of (integer) joules from the specification: [mass(grams)][length(centimeters)^2][time(seconds)^-2] * 10^-7 the "Unit Exponent" would be -7 (or 0xF9, if stored as a byte) and the "Unit" value would be 0xE121, signifying: Nibble Part Value Meaning ----- ---- ----- ------- 0 System 1 SI Linear 1 Length 2 Centimeters^2 2 Mass 1 Grams 3 Time -2 Seconds^-2 To give the resolution in e.g. hundredth of joules the "Unit Exponent" item value should have been -9. See also the examples of "Unit" values for some common units in the same chapter. However, there is a common misunderstanding about the "Unit Exponent" value encoding, where it is assumed to be stored the same as nibbles in "Unit" item. This is most likely due to the specification being a bit vague and overloading the term "unit exponent". This also was and still is proliferated by the official "HID Descriptor Tool", which makes this mistake and stores "Unit Exponent" as such. This format is also mentioned in books such as "USB Complete" and in Microsoft's hardware design guides. As a result many devices currently on the market use this encoding and so the driver should support them. Signed-off-by: Nikolai Kondrashov <[email protected]> Acked-by: Benjamin Tissoires <[email protected]> Signed-off-by: Jiri Kosina <[email protected]>
Overhaul of MAINTAINERS for Tegra. This adds Thierry as a Tegra core maintainer, and adds specific entries for most individual Tegra-specific device drivers, pointing at relevant people. The tegradrm section is updated to be Supported since Thierry is now employed to work on this. Signed-off-by: Stephen Warren <[email protected]> Acked-by: Thierry Reding <[email protected]>
We need this to work around a corruption when the boot kernel image loads the hibernated kernel image from swap on Haswell systems - somehow not everything is properly shut off. This is just the prep work, the next patch will implement the actual workaround. Signed-off-by: Ben Widawsky <[email protected]> [danvet: Add a commit message suitable for -fixes and add cc: stable] Cc: [email protected] Signed-off-by: Daniel Vetter <[email protected]>
Once the machine gets to a certain point in the suspend process, we expect the GPU to be idle. If it is not, we might corrupt memory. Empirically (with an early version of this patch) we have seen this is not the case. We cannot currently explain why the latent GPU writes occur. In the technical sense, this patch is a workaround in that we have an issue we can't explain, and the patch indirectly solves the issue. However, it's really better than a workaround because we understand why it works, and it really should be a safe thing to do in all cases. The noticeable effect other than the debug messages would be an increase in the suspend time. I have not measure how expensive it actually is. I think it would be good to spend further time to root cause why we're seeing these latent writes, but it shouldn't preclude preventing the fallout. NOTE: It should be safe (and makes some sense IMO) to also keep the VALID bit unset on resume when we clear_range(). I've opted not to do this as properly clearing those bits at some later point would be extra work. v2: Fix bugzilla link Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=65496 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=59321 Tested-by: Takashi Iwai <[email protected]> Tested-by: Paulo Zanoni <[email protected]> Signed-off-by: Ben Widawsky <[email protected]> Tested-By: Todd Previte <[email protected]> Cc: [email protected] Signed-off-by: Daniel Vetter <[email protected]>
I plan to stay with the Rockchip SoCs for the foreseable future and hope to expand its support along the way. Signed-off-by: Heiko Stuebner <[email protected]>
We can't be holding tree locks while we try to start a transaction, we will deadlock. Thanks, Reported-by: Sage Weil <[email protected]> Signed-off-by: Josef Bacik <[email protected]> Signed-off-by: Chris Mason <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.