- 29 May, 2008 10 commits
-
-
Ilpo Järvinen authored
I tried to group recovery related fields nearby (non-CA_Open related variables, to be more accurate) so that one to three cachelines would not be necessary in CA_Open. These are now contiguously deployed: struct sk_buff_head out_of_order_queue; /* 1968 80 */ /* --- cacheline 32 boundary (2048 bytes) --- */ struct tcp_sack_block duplicate_sack[1]; /* 2048 8 */ struct tcp_sack_block selective_acks[4]; /* 2056 32 */ struct tcp_sack_block recv_sack_cache[4]; /* 2088 32 */ /* --- cacheline 33 boundary (2112 bytes) was 8 bytes ago --- */ struct sk_buff * highest_sack; /* 2120 8 */ int lost_cnt_hint; /* 2128 4 */ int retransmit_cnt_hint; /* 2132 4 */ u32 lost_retrans_low; /* 2136 4 */ u8 reordering; /* 2140 1 */ u8 keepalive_probes; /* 2141 1 */ /* XXX 2 bytes hole, try to pack */ u32 prior_ssthresh; /* 2144 4 */ u32 high_seq; /* 2148 4 */ u32 retrans_stamp; /* 2152 4 */ u32 undo_marker; /* 2156 4 */ int undo_retrans; /* 2160 4 */ u32 total_retrans; /* 2164 4 */ ...and they're then followed by URG slowpath & keepalive related variables. Head of the out_of_order_queue always needed for empty checks, if that's empty (and TCP is in CA_Open), following ~200 bytes (in 64-bit) shouldn't be necessary for anything. If only OFO queue exists but TCP is in CA_Open, selective_acks (and possibly duplicate_sack) are necessary besides the out_of_order_queue but the rest of the block again shouldn't be (ie., the other direction had losses). As the cacheline boundaries depend on many factors in the preceeding stuff, trying to align considering them doesn't make too much sense. Commented one ordering hazard. There are number of low utilized u8/16s that could be combined get 2 bytes less in total so that the hole could be made to vanish (includes at least ecn_flags, urg_data, urg_mode, frto_counter, nonagle). Signed-off-by:
Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Acked-by:
Eric Dumazet <dada1@cosmosbay.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Signed-off-by: -
Matt Carlson authored
This patch increments the version to 3.93. Signed-off-by:
Matt Carlson <mcarlson@broadcom.com> Signed-off-by:
Michael Chan <mchan@broadcom.com> Signed-off-by:
Benjamin Li <benli@broadcom.com> Signed-off-by:
-
Matt Carlson authored
This patch adds some options obtained through shared memory. Signed-off-by:
-
Matt Carlson authored
This patch added the 5785 device ID and ASIC revision to the code. Signed-off-by:
-
Matt Carlson authored
This patch introduces the libphy support. Signed-off-by:
-
Matt Carlson authored
This patch introduces code to register and unregister the tg3 mdio bus with the system. Signed-off-by:
-
Matt Carlson authored
This patch introduces the TG3_FLG3_USE_PHYLIB flag and applies it to some select places. This work makes later patches a little easier to read. Signed-off-by:
-
Matt Carlson authored
This patch applies cleanups that would otherwise clutter later patches. Signed-off-by:
-
Matt Carlson authored
This patch moves some functions towards the top of the file to avoid unnecessary function prototypes. Signed-off-by:
-
- 26 May, 2008 4 commits
-
-
Alan Cox authored
Signed-off-by:
Alan Cox <alan@redhat.com> Signed-off-by:
-
Alan Cox authored
Signed-off-by:
-
Alan Cox authored
I've pushed it down as far as I dare at this point. Someone familiar with the internal PPP semantics can probably push it further. Another step to eliminating the old BKL ioctl usage. Signed-off-by:
-
David S. Miller authored
Conflicts: drivers/net/cpmac.c net/mac80211/mlme.c
-
- 23 May, 2008 2 commits
-
-
Patrick McHardy authored
Herbert Xu points out that the use of seperate feature bits for features to be propagated to VLAN devices is going to get messy real soon. Replace the VLAN feature bits by a bitmask of feature flags to be propagated and restore the old GSO_SHIFT/MASK values. Signed-off-by:
Patrick McHardy <kaber@trash.net> Acked-by:
Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
David S. Miller authored
-
- 22 May, 2008 24 commits
-
-
Komuro authored
fmvj18x_cs: The manfid of "NextCom NC5310 rev B" is MANF_ID_FUJITSU. but this card is MBH10302 based card. use ConfigBase to detect the cardtype for this card. Signed-off-by:Komuro <komurojun-mbn@nifty.com> Signed-off-by:
Jeff Garzik <jgarzik@redhat.com>
-
Komuro authored
keep bit7,8 of XIRCREG42_SWC1 in set_multicast_list. Signed-off-by:
-
Wang Chen authored
If alloc_skb failed, the recieved packet will be dropped. Do not increase rx_bytes for dropped packet. Signed-off-by:
Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by:
-
Wang Chen authored
Use standard routine for queue purging. Signed-off-by:
-
Wang Chen authored
Use standard routine for queue purging. Signed-off-by:
-
Paul Gortmaker authored
Commit cac1f3c8 factored out the code for get_phy_id so that it could be reused in multiple places. Turns out that some of the users can be modular, so we need to export this symbol as well. Signed-off-by:
Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by:
-
Thomas Graf authored
The purpose of nla_parse_nested_compat() is to parse attributes which contain a struct followed by a stream of nested attributes. So far, it called nla_parse_nested() to parse the stream of nested attributes which was wrong, as nla_parse_nested() expects a container attribute as data which holds the attribute stream. It needs to call nla_parse() directly while pointing at the next possible alignment point after the struct in the beginning of the attribute. With this patch, I can no longer reproduce the reported leftover warnings. Signed-off-by:
Thomas Graf <tgraf@suug.ch> Acked-by:
-
Krzysztof Halasa authored
WAN: protect protocol list in hdlc.c with RTNL. Signed-off-by:
Krzysztof Hałasa <khc@pm.waw.pl> Signed-off-by:
-
Julia Lawall authored
The following code appears in the function fs_init_instance in the file drivers/net/fs_enet/fs_enet-main.c. if (fep->ops == NULL) { printk(KERN_ERR DRV_MODULE_NAME ": %s No matching ops found (%d).\n", ndev->name, fpi->fs_no); err = -EINVAL; goto err; } This code implies that at the point of err, fep->ops can be NULL, so an extra test is needed before dereferencing this value. This problem was found using the following semantic match (http://www.emn.fr/x-info/coccinelle/) // <smpl> @@ expression E, E1; identifier f; statement S1,S2,S3; @@ * if (E == NULL) { ... when != if (E == NULL) S1 else S2 when != E = E1 * E->f ... when any return ...; } else S3 // </smpl> Signed-off-by:Julia Lawall <julia@diku.dk> Signed-off-by:
-
Sreenivasa Honnur authored
- Updated version number Signed-off-by:
Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by:
Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by:
-
Sreenivasa Honnur authored
- Added napi support when MSIX is enabled. - Moved test_msi function from s2io_open to probe function. Signed-off-by:
-
Sreenivasa Honnur authored
- Move all the transmit completions to a single msi-x (alarm) vector. - Enable the continuous timer interrupt for only one transmit fifo. Signed-off-by:
Santosh Rastapur <santosh.rastapur@neterion.com> Signed-off-by:
-
Joe Perches authored
Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Francois Romieu authored
The lifespan of the device covers the request_irq .. free_irq interval. The cast of a void * pointer is not needed either. Signed-off-by:
Francois Romieu <romieu@fr.zoreil.com> Signed-off-by:
-
Huang Weiyi authored
Signed-off-by:
Huang Weiyi <weiyi.huang@gmail.com> Signed-off-by:
Bryan Wu <cooloney@kernel.org> Signed-off-by:
-
Matteo Croce authored
* Resolve some locking issues using atomic_inc/atomic_dec * move status code in cpmac_check_status * unmark the BROKEN flag in Kconfig * move code which should have been in platform code in arch/mips/ar7/platform.c * fixed an IRQ storm which lets the kernel hang * fixed a double call to netif_start_queue which causes a kernel panic * don't fail to register the PHY, works on many devices now Signed-off-by:
Matteo Croce <matteo@openwrt.org> Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Becky Bruce authored
The use of unsigned long causes the driver to fail on 32-bit systems which support 64-bit resources. Signed-off-by:
Becky Bruce <becky.bruce@freescale.com> Signed-off-by:
-
Aurelien Nephtali authored
Add support for Apple USB Ethernet Adapter. http://store.apple.com/1-800-MY-APPLE/WebObjects/AppleStore.woa/wa/RSLID?nplm=MB442Z/ASigned-off-by:
Aurelien Nephtali <aurelien@sitadelle.com> Acked-by:
Greg KH <greg@kroah.com> Cc: David Brownell <david-b@pacbell.net> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Anton Vorontsov authored
This patch adds netpoll support for the uli526x ethernet driver -- simply call the interrupt handler for polling. To do this without disable_irq()/enable_irq() pair we should fully protect the handler. Luckily, it's already using irqsave spinlock, the only unprotected place is interrupts re-enabling write. It was safe to re-enable interrupts without holding the spinlock, but with netpoll possibility now it doesn't seem so. Patch was tested using netconsole and KGDBoE. Signed-off-by:
Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by:
-
Andy Fleming authored
We were passing NULL as the device. When we actually start supporting more interesting memory configurations, this will break things, so we proactively are fixing the bug. Signed-off-by:
Andy Fleming <afleming@freescale.com> Signed-off-by:
-
Maciej W. Rozycki authored
PHYLIB was first marked as BROKEN on S390, then the enclosing menu marked as non-S390, then the two dependencies merged with the conversion to menuconfig. Reduce to non-S390. Signed-off-by:
Maciej W. Rozycki <macro@linux-mips.org> Signed-off-by:
-
Gerrit Renker authored
IPv4 would work with this driver only with static arp table entries, the patch reverts a padto introduced in commit 26a17b7b sc92031: start transmit return value bugfix The padto does not work because the driver code evaluates `len' later on and there are cases where skb->len is not updated accordingly. This was observed with ARP frames (skb->len = 42 bytes, !skb_cloned(), skb_tailroom = 84 bytes). Then in skb_pad(), the first condition is true, where skb->len is not updated. As a consequence, the driver uses 42 bytes instead of the 60 bytes, and the ARP frame never makes it onto the wire. Signed-off-by:
Gerrit Renker <gerrit@erg.abdn.ac.uk> Signed-off-by:
-
Stephen Hemminger authored
Use netdev_alloc_skb. This sets skb->dev and allows arch specific allocation. Also simplify and cleanup the alignment code. Signed-off-by:
Stephen Hemminger <shemminger@vyatta.com> Signed-off-by:
-
Tobias Diedrich authored
nv_open() resets multicast settings, call nv_set_multicast(dev) to restore them. (Maybe this should rather be moved into nv_open()) Signed-off-by:
Tobias Diedrich <ranma+kernel@tdiedrich.de> Signed-off-by:
-
