aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
* x86, uv: add Kconfig dependency on NUMA for UV systemsJack Steiner2009-04-08
| | | | | | | | | | | | | Impact: build fix Add Kconfig dependency on NUMA for enabling UV. Although it might be possible to configure non-NUMA UV systems, they are unsupported and not interesting. Much of the infrastructure for UV requires NUMA support. Signed-off-by: Jack Steiner <steiner@sgi.com> LKML-Reference: <20090403203942.GA20137@sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* x86: prevent /sys/firmware/sgi_uv from being created on non-uv systemsRuss Anderson2009-04-08
| | | | | | | | /sys/firmware/sgi_uv should only be created on uv systems. Signed-off-by: Russ Anderson <rja@sgi.com> LKML-Reference: <20090403222423.GA28546@sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* x86, UV: Fix for nodes with memory and no cpusJack Steiner2009-04-03
| | | | | | | | | Fix initialization of UV blade information for systems that have nodes with memory but no cpus. Signed-off-by: Jack Steiner <steiner@sgi.com> LKML-Reference: <20090330140111.GA18461@sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* x86, UV: system table in bios accessed after unmapRuss Anderson2009-04-03
| | | | | | | | | Use the copy of UV system table in kernel memory, not the one in bios after unmapping. Signed-off-by: Russ Anderson <rja@sgi.com> LKML-Reference: <20090330225240.GA22776@sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* x86: UV BAU messaging timeoutsCliff Wickman2009-04-03
| | | | | | | | | | | | | | | | | This patch replaces a 'nop' uv_enable_timeouts() in the UV TLB shootdown code. (somehow, long ago that function got eviscerated) If any cpu in the destination node does not get interrupted by the message and post completion in a reasonable time the hardware should respond to the sender with an error. This function enables such timeouts. Tested on the UV hardware simulator. Signed-off-by: Cliff Wickman <cpw@sgi.com> LKML-Reference: <E1LpjXU-00007e-Qh@eag09.americas.sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* x86: UV BAU and nodes with no memoryCliff Wickman2009-04-03
| | | | | | | | | | | | | | | This patch fixes BAU initialization for systems containing nodes with no memory and for systems with non-consecutive node numbers. Fixes and clarifies situations where pnode should be used instead of node id. Tested on the UV hardware simulator. Signed-off-by: Cliff Wickman <cpw@sgi.com> LKML-Reference: <E1LpjX3-00007N-12@eag09.americas.sgi.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
* Merge branch 'dma-debug' of ↵Ingo Molnar2009-04-03
|\ | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/joro/linux-2.6-iommu into x86/urgent
| * x86/dma: unify definition of pci_unmap_addr* and pci_unmap_len macrosJoerg Roedel2009-04-03
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Impact: unification of pci-dma macros and pci_32.h removal This patch unifies the definition of the pci_unmap_addr*, pci_unmap_len* and DECLARE_PCI_UNMAP* macros. This makes sense because the pci_unmap functions are no longer no-ops anymore when the kernel runs with CONFIG_DMA_API_DEBUG. Without an iommu or DMA_API_DEBUG it is a no-op on 32 bit because the dma mapping path returns a physical address and therefore the dma-api implementation has no internal state which needs to be destroyed with an unmap call. This unification also simplifies the port of x86_64 iommu drivers to 32 bit x86 and let us get rid of pci_32.h. Signed-off-by: Joerg Roedel <joerg.roedel@amd.com> Acked-by: Stephen Hemminger <shemminger@vyatta.com>
| * Merge branch 'for-linus' of ↵Linus Torvalds2009-04-03
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6 * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: Remove two unneeded exports and make two symbols static in fs/mpage.c Cleanup after commit 585d3bc06f4ca57f975a5a1f698f65a45ea66225 Trim includes of fdtable.h Don't crap into descriptor table in binfmt_som Trim includes in binfmt_elf Don't mess with descriptor table in load_elf_binary() Get rid of indirect include of fs_struct.h New helper - current_umask() check_unsafe_exec() doesn't care about signal handlers sharing New locking/refcounting for fs_struct Take fs_struct handling to new file (fs/fs_struct.c) Get rid of bumping fs_struct refcount in pivot_root(2) Kill unsharing fs_struct in __set_personality()
| | * Remove two unneeded exports and make two symbols static in fs/mpage.cDmitri Vorobiev2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 29a814d2ee0e43c2980f33f91c1311ec06c0aa35 (vfs: add hooks for ext4's delayed allocation support) exported the following functions mpage_bio_submit() __mpage_writepage() for the benefit of ext4's delayed allocation support. Since commit a1d6cc563bfdf1bf2829d3e6ce4d8b774251796b (ext4: Rework the ext4_da_writepages() function), these functions are not used by the ext4 driver anymore. However, the now unnecessary exports still remain, and this patch removes those. Moreover, these two functions can become static again. The issue was spotted by namespacecheck. Signed-off-by: Dmitri Vorobiev <dmitri.vorobiev@movial.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Cleanup after commit 585d3bc06f4ca57f975a5a1f698f65a45ea66225Al Viro2009-04-01
| | | | | | | | | | | | | | | | | | | | | fsync_bdev() export and a bunch of stubs for !CONFIG_BLOCK case had been left behind Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Trim includes of fdtable.hAl Viro2009-03-31
| | | | | | | | | | | | Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Don't crap into descriptor table in binfmt_somAl Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | Same story as in binfmt_elf, except that in binfmt_som we actually forget to close the sucker. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Trim includes in binfmt_elfAl Viro2009-03-31
| | | | | | | | | | | | Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Don't mess with descriptor table in load_elf_binary()Al Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | | | | ... since we don't tell anyone which descriptor does the file get. We used to, but only in case of ELF binary with a.out loader and that stuff has been gone for a while. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Get rid of indirect include of fs_struct.hAl Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | | | | Don't pull it in sched.h; very few files actually need it and those can include directly. sched.h itself only needs forward declaration of struct fs_struct; Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * New helper - current_umask()Al Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | current->fs->umask is what most of fs_struct users are doing. Put that into a helper function. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * check_unsafe_exec() doesn't care about signal handlers sharingAl Viro2009-03-31
| | | | | | | | | | | | | | | | | | ... since we'll unshare sighand anyway Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * New locking/refcounting for fs_structAl Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * all changes of current->fs are done under task_lock and write_lock of old fs->lock * refcount is not atomic anymore (same protection) * its decrements are done when removing reference from current; at the same time we decide whether to free it. * put_fs_struct() is gone * new field - ->in_exec. Set by check_unsafe_exec() if we are trying to do execve() and only subthreads share fs_struct. Cleared when finishing exec (success and failure alike). Makes CLONE_FS fail with -EAGAIN if set. * check_unsafe_exec() may fail with -EAGAIN if another execve() from subthread is in progress. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Take fs_struct handling to new file (fs/fs_struct.c)Al Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pure code move; two new helper functions for nfsd and daemonize (unshare_fs_struct() and daemonize_fs_struct() resp.; for now - the same code as used to be in callers). unshare_fs_struct() exported (for nfsd, as copy_fs_struct()/exit_fs() used to be), copy_fs_struct() and exit_fs() don't need exports anymore. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Get rid of bumping fs_struct refcount in pivot_root(2)Al Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | | | | Not because execve races with _that_ are serious - we really need a situation when final drop of fs_struct refcount is done by something that used to have it as current->fs. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| | * Kill unsharing fs_struct in __set_personality()Al Viro2009-03-31
| | | | | | | | | | | | | | | | | | | | | That's a rudiment of altroot support. I.e. it should've been buried a long time ago. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
| * | Merge branch 'drm-linus' of ↵Linus Torvalds2009-04-03
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6 * 'drm-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6: (21 commits) drm/radeon: load the right microcode on rs780 drm: remove unused "can_grow" parameter from drm_crtc_helper_initial_config drm: fix EDID backward compat check drm: sync the mode validation for INTERLACE/DBLSCAN drm: fix typo in edid vendor parsing. DRM: drm_crtc_helper.h doesn't actually need i2c.h drm: fix missing inline function on 32-bit powerpc. drm: Use pgprot_writecombine in GEM GTT mapping to get the right bits for !PAT. drm/i915: Add a spinlock to protect the active_list drm/i915: Fix SDVO TV support drm/i915: Fix SDVO CREATE_PREFERRED_INPUT_TIMING command drm/i915: Fix error in SDVO DTD and modeline convert drm/i915: Fix SDVO command debug function drm/i915: fix TV mode setting in property change drm/i915: only set TV mode when any property changed drm/i915: clean up udelay usage drm/i915: add VGA hotplug support for 945+ drm/i915: correctly set IGD device's gtt size for KMS. drm/i915: avoid hanging on to a stale pointer to raw_edid. drm/i915: check for -EINVAL from vm_insert_pfn ...
| | * | drm/radeon: load the right microcode on rs780Alex Deucher2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Copy/paste error. The RV670 microcode should work ok, so it's not a show stopper. Signed-off-by: Alex Deucher <alexdeucher@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | Merge branch 'drm-intel-next' of ../anholt-2.6 into drm-linusDave Airlie2009-04-02
| | |\ \
| | | * | drm: Use pgprot_writecombine in GEM GTT mapping to get the right bits for !PAT.Jesse Barnes2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Otherwise, the PAGE_CACHE_WC would end up getting us a UC-only mapping, and the write performance of GTT maps dropped 10x. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> [anholt: cleaned up unused var] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: Add a spinlock to protect the active_listCarl Worth2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a baby-step in the direction of having finer-grained locking than the struct_mutex. Specifically, this will enable new debugging code to read the active list for printing out GPU state when the GPU is wedged, (while the struct_mutex is held, of course). Signed-off-by: Carl Worth <cworth@cworth.org> [anholt: indentation fix] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: Fix SDVO TV supportZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This brings SDVO TV support from 2D driver, including origin fix f1ca56e17d0 and later fix 2fcf4fcccfe. Also fix wrong modeline definitions for SDVO TV. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: Fix SDVO CREATE_PREFERRED_INPUT_TIMING commandZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This brings fix commit acde0ef683 from 2D driver. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: Fix error in SDVO DTD and modeline convertZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This brings fix commit 62c0c2f5549a from 2D driver. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: Fix SDVO command debug functionZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix compile error of intel_sdvo_debug_response(), and explicit use KERN_DEBUG for printk. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: fix TV mode setting in property changeZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Only set TV DAC in property change seems doesn't work, we have to setup whole crtc pipe which assigned to TV alone. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> [anholt: Note that this should also fix the oops at startup with new 2D] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: only set TV mode when any property changedZhenyu Wang2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If there's no real property change, don't need to set TV mode again. Signed-off-by: Zhenyu Wang <zhenyu.z.wang@intel.com> [anholt: checkpatch.pl fix] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: clean up udelay usageArjan van de Ven2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | udelay() of 20 milliseconds really ought to just use mdelay(), that avoids the various wrap scenarios and also is more readable Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: add VGA hotplug support for 945+Jesse Barnes2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add VGA port hotplug detection to the i915 driver. When KMS is enabled, plugging in or removing a VGA cable from the VGA connector will generate a uevent, which indicates to userspace that it should re-probe outputs on this device (to determine modes, etc.). Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> [anholt: dropped extra PORT_HOTPLUG_STAT clear with ack from jbarnes] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: correctly set IGD device's gtt size for KMS.Shaohua Li2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | IGD device only has last 1 page used by GTT. This should match the AGP gart code. Signed-off-by: Shaohua Li <shaohua.li@intel.com> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: avoid hanging on to a stale pointer to raw_edid.Ma Ling2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | drm_get_edid will store edid into raw_edid, so when freeing edid memory, at the same time clean raw_edid pointer. Signed-off-by: Ma Ling <ling.ma@intel.com> [anholt: Note that raw_edid is not currently used anywhere] Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: check for -EINVAL from vm_insert_pfnJesse Barnes2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Indicates something is wrong with the mapping; and apparently triggers in current kernels. Signed-off-by: Jesse Barnes <jbarnes@virtuosugeek.org> Signed-off-by: Eric Anholt <eric@anholt.net>
| | | * | drm/i915: fix up tiling/fence reg setup on i8xx class hwDaniel Vetter2009-04-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes all the tiling problems with the 2d ddx. glxgears still doesn't work. Changes: - fix a copy&paste error in i8xx fence reg setup. It resulted in an at most a 512KB offset of the fence reg window, so was only visible sometimes. - add tests for stride and object size constrains (also for i915 and 1965 class hw). Userspace seems to have an of-by-one bug there, which changes the fence size by at most 512KB due to an overflow. - because i8xx hw is quite old (and therefore not as well-tested) I left 2 debug WARN_ONs in the i8xx fence reg setup code to hopefully catch any further overflows in the bit-fields. Lastly there's one small change to make the alignment checks more consistent. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=20289 Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Eric Anholt <eric@anholt.net>
| | * | | drm: remove unused "can_grow" parameter from drm_crtc_helper_initial_configJesse Barnes2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cleanup some leftovers from the X port. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | | drm: fix EDID backward compat checkJesse Barnes2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | EDIDs should be backward compatible, so don't bail if we see a version of 3 (which is out there now) and print a message if we see something newer, but allow it to be parsed. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | | drm: sync the mode validation for INTERLACE/DBLSCANyakui_zhao2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Check whether the INTERLACE/DBLSCAN is supported by output device. If not, the mode containing the flag of INTERLACE/DBLSCAN will be marked as unsupported. Signed-off-by: Zhao Yakui <yakui.zhao@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | | drm: fix typo in edid vendor parsing.Dave Airlie2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Should be, edid_vendor[2] = (edid->mfg_id[1] & 0x1f) + '@'; Since vendor ID has only two bytes only, I am somewhat surprised why gcc doesn't complain this. Reported-by: Guo, Chaohong <chaohong.guo@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | | DRM: drm_crtc_helper.h doesn't actually need i2c.hJean Delvare2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove an include that isn't actually needed to prevent needless rebuilds. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
| | * | | drm: fix missing inline function on 32-bit powerpc.Dave Airlie2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The readq/writeq really need to be static inline on the arches which don't provide them. Reported-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
| * | | | Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6Linus Torvalds2009-04-03
| |\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (54 commits) glge: remove unused #include <version.h> dnet: remove unused #include <version.h> tcp: miscounts due to tcp_fragment pcount reset tcp: add helper for counter tweaking due mid-wq change hso: fix for the 'invalid frame length' messages hso: fix for crash when unplugging the device fsl_pq_mdio: Fix compile failure fsl_pq_mdio: Revive UCC MDIO support ucc_geth: Pass proper device to DMA routines, otherwise oops happens i.MX31: Fixing cs89x0 network building to i.MX31ADS tc35815: Fix build error if NAPI enabled hso: add Vendor/Product ID's for new devices ucc_geth: Remove unused header gianfar: Remove unused header kaweth: Fix locking to be SMP-safe net: allow multiple dev per napi with GRO r8169: reset IntrStatus after chip reset ixgbe: Fix potential memory leak/driver panic issue while setting up Tx & Rx ring parameters ixgbe: fix ethtool -A|a behavior ixgbe: Patch to fix driver panic while freeing up tx & rx resources ...
| | * | | | glge: remove unused #include <version.h>Huang Weiyi2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove unused #include <version.h> in drivers/net/qlge/qlge_ethtool. Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| | * | | | dnet: remove unused #include <version.h>Huang Weiyi2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove unused #include <version.h> in drivers/net/dnet.c. Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| | * | | | tcp: miscounts due to tcp_fragment pcount resetIlpo Järvinen2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It seems that trivial reset of pcount to one was not sufficient in tcp_retransmit_skb. Multiple counters experience a positive miscount when skb's pcount gets lowered without the necessary adjustments (depending on skb's sacked bits which exactly), at worst a packets_out miscount can crash at RTO if the write queue is empty! Triggering this requires mss change, so bidir tcp or mtu probe or like. Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de> Tested-by: Uwe Bugla <uwe.bugla@gmx.de> Signed-off-by: David S. Miller <davem@davemloft.net>
| | * | | | tcp: add helper for counter tweaking due mid-wq changeIlpo Järvinen2009-04-02
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We need full-scale adjustment to fix a TCP miscount in the next patch, so just move it into a helper and call for that from the other places. Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.2 (git 2.25.0) at 2024-12-03 06:15:01 -0500