From: Jeff Kirsher
> Sent: 25 November 2015 10:47
> On Tue, 2015-11-24 at 16:43 -0800, Joe Perches wrote:
> > On Tue, 2015-11-24 at 16:04 -0800, Jeff Kirsher wrote:
> > > From: Helin Zhang
> > >
> > > Pointer of type void * shouldn't be used in arithmetic, which may
> > >
From: Marcelo Ricardo Leitner
> Sent: 30 November 2015 16:33
> Dmitry Vyukov reported that the user could trigger a kernel warning by
> using a large len value for getsockopt SCTP_GET_LOCAL_ADDRS, as that
> value directly affects the value used as a kmalloc() parameter.
>
> This patch thus
From: Will Deacon
> Sent: 18 November 2015 10:14
> On Tue, Nov 17, 2015 at 08:17:17PM +0100, Arnd Bergmann wrote:
> > On Tuesday 17 November 2015 17:12:37 Will Deacon wrote:
> > > On Tue, Nov 17, 2015 at 06:03:40PM +0100, Arnd Bergmann wrote:
> > > > On Tuesday 17 November 2015 16:44:53 Will
From: Maciej Zenczykowski
> Sent: 18 November 2015 03:57
> I don't know what the right fix is...
>
> However, speaking as an end user with laptops on wifi and/or home
> gateways on dialup connections where the IP address occasionally (or
> constantly) changes, I find it very frustrating that by
From: Arnd Bergmann
> Sent: 17 November 2015 19:17
> On Tuesday 17 November 2015 17:12:37 Will Deacon wrote:
> > On Tue, Nov 17, 2015 at 06:03:40PM +0100, Arnd Bergmann wrote:
> > > On Tuesday 17 November 2015 16:44:53 Will Deacon wrote:
> > > > > 8<
> > > > > Subject: ARM64: make
From: Florian Westphal
> Sent: 21 November 2015 16:56
> > +struct xt_cgroup_info_v1 {
> > + __u8has_path;
> > + __u8has_classid;
> > + __u8invert_path;
> > + __u8invert_classid;
> > + charpath[PATH_MAX];
> > + __u32
From: Yuval Mintz
> Sent: 22 November 2015 20:19
> This might be a dumb question, but I recently touched this
> and felt like I'm missing something basic -
>
> NAPI is being scheduled from soft-interrupt contex, and it
> has a ~strict quota for handling Rx packets [even though we're
> allowing
From: Eric Dumazet
> Sent: 09 June 2016 22:17
> On Thu, 2016-06-09 at 23:50 +0300, Saeed Mahameed wrote:
> > From: Matthew Finlay
>
>
> > diff --git a/net/socket.c b/net/socket.c
> > index a1bd161..67de200 100644
> > --- a/net/socket.c
> > +++ b/net/socket.c
> > @@ -382,6
From: Tariq Toukan
> Sent: 26 May 2016 10:39
...
> I am aware that clearing the stats structure might be redundant today,
> as the function is called only within mlx4_en_open, but we might want to
> call the function in other flows in the future.
My personal view is that stats should never be
From: Jakub Sitnicki
> Sent: 22 June 2016 12:34
...
> > - a = attrs[TCP_METRICS_ATTR_ADDR_IPV4];
> > - if (a) {
> > + if ((a = attrs[TCP_METRICS_ATTR_ADDR_IPV4])) {
>
> Copy the pointer inside the branch?
>
> Same gain on indentation while keeping checkpatch happy.
Or as below (hacked
From: Behalf Of David Ahern
> Sent: 17 June 2016 00:24
> VRF driver needs access to ip6_route_get_saddr code. Since it does
> little beyond ipv6_dev_get_saddr and ipv6_dev_get_saddr is already
> exported for modules move ip6_route_get_saddr to the header as an
> inline.
>
> Code move only; no
From: Of Phil Sutter
> Sent: 21 June 2016 17:19
> By combining the attribute extraction and check for existence, the
> additional indentation level in the 'else' clause can be avoided.
>
> In addition to that, common actions for 'daddr' are combined since the
> function returns if neither of the
From: Arnd Bergmann
> On Wednesday, June 15, 2016 5:10:51 PM CEST Jes Sorensen wrote:
> >
> > Arnd,
> >
> > rtlwifi and rtl8xxxu are two distinct drivers managed by different
> > people. I'd be really nice if you could split this into a per driver
> > patch.
> >
> > That said, the use of char in
From: Phil Sutter
> Sent: 23 June 2016 18:34
>
> By directly comparing the value of both unsigned variables, casting to
> signed becomes unnecessary.
>
> This also fixes for compiling with older versions of gcc (at least
> <=3.4.6) which emit the following warning:
>
> | ifstat.c: In function
From: Phil Sutter
> Sent: 23 June 2016 18:34
>
> This is v3 of my C99-style initializer related patch series.
...
It would be interesting to know how this affect the kernel code size?
While gcc will generate a memset() call for 'struct foo = {0}' if you
initialise some members it might generate
From: George Spelvin
> Sent: 10 February 2016 14:44
...
> > I think the fastest loop is:
> > 10: adcq0(%rdi,%rcx,8),%rax
> > inc %rcx
> > jnz 10b
> > That loop looks like it will have no overhead on recent cpu.
>
> Well, it should execute at 1 instruction/cycle.
I presume you
From: Francois Romieu
> Sent: 06 February 2016 10:59
> > +void mpls_stats_inc_outucastpkts(struct net_device *dev,
> > +const struct sk_buff *skb)
> > +{
> > + struct mpls_dev *mdev;
> > + struct inet6_dev *in6dev;
>
> Nit: the scope can be reduced for both
From: Larry Finger
> Sent: 03 February 2016 19:45
...
> The performance will depend on where you satisfy the condition. All switch
> cases
> have the same execution time, but in the if .. else if .. else form, the
> earlier
> tests execute more quickly. I'm not sure that one can make any blanket
From: Tom Herbert
> Sent: 03 February 2016 19:19
...
> + /* Main loop */
> +50: adcq0*8(%rdi),%rax
> + adcq1*8(%rdi),%rax
> + adcq2*8(%rdi),%rax
> + adcq3*8(%rdi),%rax
> + adcq4*8(%rdi),%rax
> + adcq5*8(%rdi),%rax
> + adcq6*8(%rdi),%rax
> +
From: George Spelvin [mailto:li...@horizon.com]
> Sent: 08 February 2016 20:13
> David Laight wrote:
> > I'd need convincing that unrolling the loop like that gives any significant
> > gain.
> > You have a dependency chain on the carry flag so have delays between the
>
From: Grygorii Strashko
> Sent: 09 February 2016 13:58
> From: Arnd Bergmann
>
> The commit 899077791403 ("netcp: try to reduce type confusion in descriptors")
> introduces a regression in Kernel 4.5-rc1 and it breaks
> get/set_pad_info() functionality.
>
> The TI NETCP driver
From: Karicheri, Muralidharan
> Sent: 09 February 2016 16:10
...
> >In reality the 'pad' fields ought to be renamed - since they aren't pads.
> >Perhaps they should be a union?
> No. At the end of the descriptor, host software can add scratchpad which is
> not modified by the hardware, but is
From: Karicheri, Muralidharan
> Sent: 09 February 2016 16:19
> >...
> >> >In reality the 'pad' fields ought to be renamed - since they aren't pads.
> >> >Perhaps they should be a union?
> >
> >> No. At the end of the descriptor, host software can add scratchpad
> >> which is not modified by the
From: Tom Herbert
...
> > If nothing else reducing the size of this main loop may be desirable.
> > I know the newer x86 is supposed to have a loop buffer so that it can
> > basically loop on already decoded instructions. Normally it is only
> > something like 64 or 128 bytes in size though. You
From: Kees Cook
> Sent: 04 February 2016 21:01
> Some callers of strtobool were passing a pointer to unterminated strings.
> In preparation of adding multi-character processing to kstrtobool, update
> the callers to not pass single-character pointers, and switch to using the
> new
From: Ingo Molnar
...
> As Linus noticed, data lookup tables are the intelligent solution: if you
> manage
> to offload the logic into arithmetics and not affect the control flow then
> that's
> a big win. The inherent branching will be hidden by executing on massively
> parallel arithmetics
From: George Spelvin
> Sent: 10 February 2016 00:54
> To: David Laight; linux-ker...@vger.kernel.org; li...@horizon.com;
> netdev@vger.kernel.org;
> David Laight wrote:
> > Since adcx and adox must execute in parallel I clearly need to re-remember
> > how dependencies ag
From: Larry Finger
> Sent: 04 February 2016 15:44
> On 02/04/2016 03:48 AM, David Laight wrote:
> > From: Larry Finger
> >> Sent: 03 February 2016 19:45
> > ...
> >> The performance will depend on where you satisfy the condition. All switch
> >
From: 'Marcelo Ricardo Leitner'
> Sent: 28 January 2016 20:56
> On Thu, Jan 28, 2016 at 05:30:24PM +, David Laight wrote:
> > From: 'Marcelo Ricardo Leitner'
> > > Sent: 28 January 2016 15:53
> > > On Thu, Jan 28, 2016 at 01:51:02PM +, David Laight wrote:
.
> From: 'Marcelo Ricardo Leitner' [mailto:marcelo.leit...@gmail.com]
> Sent: 28 January 2016 20:56
...
> > > But yes, agreed, MSG_MORE is at least a welcomed compliment here,
> > > specially for applications generating a train of chunks. Will put that in
> > > my ToDo here, thanks.
> >
> > I've
From: 'Marcelo Ricardo Leitner'
> Sent: 28 January 2016 20:56
...
> > > > I did wonder whether the queued data could actually be picked up
> > > > be a Heartbeat chunk that is probing a different remote address
> > > > (which would be bad news).
> > >
> > > I don't follow. You mean if a heartbeat
From: Arnd Bergmann
> Sent: 27 January 2016 14:05
> The moxart ethernet driver confuses coherent DMA buffers with
> MMIO registers.
>
> moxart_ether.c: In function 'moxart_mac_setup_desc_ring':
> moxart_ether.c:146:428: error: passing argument 1 of '__fswab32' makes
> integer from pointer
From: Lucas Tanure
> Sent: 30 January 2016 13:18
> As suggested by checkpatch.pl:
> CHECK: Prefer kernel type 'uX' over 'uintX_t'
One might ask why?
The kernel types are older, but the uintX_t ones are now part
of the C standard.
Writing header files (eg for ioctl buffers) that have to be
parsed
From: Florian Westphal
> Sent: 30 January 2016 21:30
> Weidong Wang wrote:
> > In the 'for(...) {}', the *bucket alwasy < net->ct.htable_size,
> > so remove the check
> > @@ -1383,14 +1383,12 @@ get_next_corpse(struct net *net, int (*iter)(struct
> > nf_conn *i, void
From: David Miller
> Sent: 31 January 2016 03:45
> To: alexander.du...@gmail.com
...
> > I would really prefer to keep the pages DMA aligned, and the skb->data
> > IP aligned. If nothing else it has the advantage of actually having
> > the proper alignment on all the headers if I only pull the
From: Linus Torvalds
> Sent: 02 February 2016 20:45
> On Tue, Feb 2, 2016 at 12:32 PM, Hannes Frederic Sowa
> wrote:
> >
> > Unfortunately we never transfer a scm_cookie via the skbs but merely use it
> > to initialize unix_skb_parms structure in skb->cb and destroy it
From: Byeoungwook Kim
> Sent: 03 February 2016 02:00
> Conditional codes in rtl_addr_delay() were improved in readability and
> performance by using switch codes.
I'd like to see the performance data :-)
> diff --git a/drivers/net/wireless/realtek/rtlwifi/core.c
>
From: 'Marcelo Ricardo Leitner'
> Sent: 28 January 2016 15:53
> On Thu, Jan 28, 2016 at 01:51:02PM +, David Laight wrote:
> > From: Marcelo Ricardo Leitner
> > > Sent: 27 January 2016 17:07
> > > This patchset is merely a RFC for the moment. There are some
>
From: Alexander Duyck
> Sent: 01 February 2016 18:18
> >> 1) Unaligned accesses
> >
> > Remember the even if you do a 'realignment copy' of the IP header,
> > at some point the 'userdata' of the packet has to be accessed.
> > Mostly this will be with memcpy() and you want that copy to be aligned.
From: Paul Burton
> Sent: 03 February 2016 12:03
> Resets of the EG20T MAC on the MIPS Boston development board take longer
> than the 1000 loops that pch_gbe_wait_clr_bit was performing. Bump up
> the number of loops.
...
> diff --git a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c
>
From: Herbert Xu
> Sent: 03 February 2016 12:21
> On Wed, Feb 03, 2016 at 12:36:21PM +0100, Hannes Frederic Sowa wrote:
> >
> > Agreed that it feels like a hack, but a rather simple one. I would
> > consider this to be just a performance improvement. We certainly need
> > a slow-path when virtio
From: Marcelo Ricardo Leitner
> Sent: 27 January 2016 17:07
> This patchset is merely a RFC for the moment. There are some
> controversial points that I'd like to discuss before actually proposing
> the patches.
You also need to look at how a 'user' can actually get SCTP to
merge data chunks in
From: Robert Richter
> Sent: 18 February 2016 12:39
> From: Robert Richter
>
> Looks like the :1 notation was accidentally introduced (this still
> uses 1 byte per flag). Using bool instead, which is the common use.
>
> Signed-off-by: Robert Richter
>
From: Joe Perches
> Sent: 18 February 2016 04:59
> On Wed, 2016-02-17 at 19:38 -0800, Jeff Kirsher wrote:
> > From: Mitch Williams
> >
> > Mr. Spock would certainly raise an eyebrow to see us using bitwise
> > operators, when we should clearly be relying on logic.
From: Guillaume Nault
> Sent: 26 February 2016 17:46
>
> ppp_read() and ppp_poll() can be called concurrently with ppp_ioctl().
> In this case, ppp_ioctl() might call ppp_ccp_closed(), which may update
> ppp->flags while ppp_read() or ppp_poll() is reading it.
> The update done by
From: David Miller
> Sent: 23 February 2016 18:25
>
> From: Jesse Gross
> Date: Tue, 23 Feb 2016 09:31:09 -0800
>
> > Most OSs (including Linux with connected TCP sockets) use non-zero IP
> > IDs so requiring this would effectively disable GRO.
>
> +1
>
> Any OS that wants
From: Eric Dumazet
> Sent: 29 January 2016 22:29
...
> On modern intel cpus, this does not matter at all, sure. It took a while
> before "rep movsb" finally did the right thing.
Unfortunately memcpy_to_io() etc now map to 'rep movsb', and that can
only be optimisied for cached addresses.
So
> > + /* ignore optional '0x' prefix */
> > + if ((slen > 2) && (
Unnecessary ().
> > + (0 == memcmp(s, "0x", 2)
> > + || (0 == memcmp(s, "0X", 2) {
A-about-F comparisons.
> memcmp() is a really poor tool for comparing strings. You should use
>
From: Alexander Potapenko
> Sent: 15 March 2016 09:04
> According to IEEE Std 1003.1, 2013, sending data to a SOCK_SEQPACKET
> socketpair with MSG_NOSIGNAL flag set must result in a SIGPIPE if the
> socket is no longer connected.
...
> Without the below patch the behavior is as follows:
>
> $
From: Joe Perches
> Sent: 08 March 2016 23:26
...
> > +
> > + if (offset & 1)
> > + sum = (sum << 24) + (sum >> 8);
>
> Maybe use ror32(sum, 8);
>
> or maybe something like:
>
> {
> u32 sum;
>
> /* rotated csum2 of odd offset will be the right checksum */
> if
From: Amitkumar Karwar
> Sent: 05 April 2016 06:48
...
> Our one time allocated 64k buffer read from firmware contains multiple data
> chunks. We have a feature
> called single port aggregation in which firmware attaches an aggregated
> buffer to single port. So
> sometimes a single data chunk
From: Petri Gynther
> Sent: 05 April 2016 01:10
...
> 2. Readability: Add parentheses around nr_frags + 1.
...
> - if (ring->free_bds <= nr_frags + 1) {
...
> + if (ring->free_bds <= (nr_frags + 1)) {
The extra () are not needed and do not improve readability.
David
From: Jiri Pirko
> Sent: 22 March 2016 09:02
> Rename hlist.h to list.h while adding it to be aligned with kernel
...
> include/hlist.h | 56
> include/list.h | 112
>
I'm not sure where this gets installed,
From: Alexander Duyck
...
> One thought I had is that we may want to look into making an inline
> function that we can call for compile-time defined lengths less than
> 64. Maybe call it something like __csum_partial and we could then use
> that in place of csum_partial for all those headers
From: Alexander Duyck
...
> >> So the loop:
> >> 10: addc %rax,(%rdx,%rcx,8)
> >> inc %rcx
> >> jnz 10b
> >> could easily be as fast as anything that doesn't use the 'new'
> >> instructions that use the overflow flag.
> >> That loop might be measurable faster for aligned
From: Linus Torvalds
> Sent: 03 March 2016 18:44
>
> On Thu, Mar 3, 2016 at 8:12 AM, David Laight <david.lai...@aculab.com> wrote:
> >
> > Did you try the asm loop that used 'leax %rcx..., jcxz... jmps..'
> > without any unrolling?
>
> Is that actually su
From: John Crispin
> Sent: 03 March 2016 20:03
...
> +/* ugly macro hack to make sure hw_stats and ethtool strings are consistent
> */
> +#define MTK_STAT_REG_DECLARE \
> + _FE(tx_bytes) \
> + _FE(tx_packets) \
> + _FE(tx_skip)
From: Tom Herbert
> Sent: 02 March 2016 22:19
...
> + /* Main loop using 64byte blocks */
> + for (; len > 64; len -= 64, buff += 64) {
> + asm("addq 0*8(%[src]),%[res]\n\t"
> + "adcq 1*8(%[src]),%[res]\n\t"
> + "adcq 2*8(%[src]),%[res]\n\t"
> +
From: Alexander Duyck
...
> Actually probably the easiest way to go on x86 is to just replace the
> use of len with (len >> 6) and use decl or incl instead of addl or
> subl, and lea instead of addq for the buff address. None of those
> instructions effect the carry flag as this is how such
From: Marcelo Ricardo Leitner
> Sent: 29 March 2016 14:42
>
> Currently on high rate SCTP streams the heartbeat timer refresh can
> consume quite a lot of resources as timer updates are costly and it
> contains a random factor, which a) is also costly and b) invalidates
> mod_timer() optimization
From: Marcelo Ricardo Leitner
> Sent: 30 March 2016 13:13
> Em 30-03-2016 06:37, David Laight escreveu:
> > From: Marcelo Ricardo Leitner
> >> Sent: 29 March 2016 14:42
> >>
> >> Currently on high rate SCTP streams the heartbeat timer refresh can
> &g
From: Johannes Berg
> Sent: 27 April 2016 10:44
> On Wed, 2016-04-27 at 12:33 +0300, Petko Manolov wrote:
> >
> > Your guess turned out to be not so wild.;)All Pegasus devices are
> > configured (by the driver) to append CRC at the end of each RX
> > packet.However, the driver reports packet
From: Jesper Dangaard Brouer
> Sent: 26 April 2016 17:27
> Make compiling samples/bpf more user friendly, by detecting if LLVM
> compiler tool 'llc' is available, and also detect if the 'bpf' target
> is available in this version of LLVM.
...
> diff --git a/samples/bpf/Makefile
From: Alexei Starovoitov
> Sent: 26 April 2016 15:35
> On Tue, Apr 26, 2016 at 01:09:32PM +0200, Jesper Dangaard Brouer wrote:
> > It is not intuitive that 'make' must be run from the top level
> > directory with argument "samples/bpf/" to compile these eBPF samples.
> >
> > Introduce a kbuild
From: Saeed Mahameed
> Sent: 22 April 2016 20:01
> Redesign the statistics in the driver:
> 1. Move counters to a separate file (en_stats.h).
> 2. Remove unnecessary dependencies between stats and strings.
> 3. Use counter descriptors which hold a name and offset for each counter,
>and will be
From: Heinrich Schuchardt
> Sent: 17 May 2016 21:01
> In the statement
> assert(priv || priv->ae_handle);
> the right side of || is only evaluated if priv is null.
>
> Signed-off-by: Heinrich Schuchardt
> ---
> drivers/net/ethernet/hisilicon/hns/hns_ethtool.c | 8
From: Jesper Dangaard Brouer
> Sent: 20 May 2016 18:50
...
> If would be cool if you could run a test with removed busylock and
> allow HTB to bulk dequeue.
(Without having looked )
Could you have two queues and separate queue and dequeue locks.
The enqueue code would acquire the enqueue
From: netdev-ow...@vger.kernel.org Akinobu Mita
> Sent: 14 May 2016 06:56
> This increases TX timeout period from one second to 5 seconds which is
> the default value if the driver doesn't explicitly set
> net_device->watchdog_timeo.
>
> The one second timeout is too short for W5100 with SPI
From: Martin Willi
> Sent: 03 May 2016 07:53
> While wiphys can be moved into network namespaces over nl80211, the
> creation and removal of hwsim radios is currently limited to the initial
> namespace. This patch allows management of namespaced radios from the
> owning namespace by setting
From: Ben Hutchings
> Sent: 09 May 2016 01:17
> On Sun, 2016-05-08 at 13:55 -0700, Shrikrishna Khare wrote:
> >
> > On Sat, 7 May 2016, Ben Hutchings wrote:
> >
> > > On Fri, 2016-05-06 at 16:12 -0700, Shrikrishna Khare wrote:
> > > [...]
> > > > +static int
> > > > +vmxnet3_set_coalesce(struct
From: Jiri Pirko
> Sent: 15 April 2016 09:53
..
> >> @@ -106,10 +96,9 @@ static int mlxsw_sp_port_pb_init(struct mlxsw_sp_port
> >> *mlxsw_sp_port)
> >>mlxsw_reg_pbmc_pack(pbmc_pl, mlxsw_sp_port->local_port,
> >>0x, 0x / 2);
> >>for (i = 0; i <
From: Joe Stringer
> Sent: 13 April 2016 19:10
> This is the IPv6 equivalent of commit 8282f27449bf ("inet: frag: Always
> orphan skbs inside ip_defrag()").
>
> Prior to commit 029f7f3b8701 ("netfilter: ipv6: nf_defrag: avoid/free
> clone operations"), ipv6 fragments sent to nf_ct_frag6_gather()
From: Jiri Pirko
> Sent: 14 April 2016 17:19
> From: Jiri Pirko
>
> Structs are in arrays so use array index as pool/tc/prio index. With
> that, there is need to maintain separate arrays for ingress and egress.
...
> +static const u16 mlxsw_sp_pbs[] = {
> + 2 *
From: Eric Dumazet
> Sent: 19 April 2016 00:18
...
> MSG_EOR should not depend on SKBTX_ANY_TSTAMP
>
> Really, simply using send(fd, ..., len, MSG_EOR) should instruct TCP to
> mark the cooked skb as a non candidate for future coalescing.
Isn't that very similar to the inverse of MSG_MORE?
Or a
From: Arvind Yadav
> Sent: 23 July 2016 19:06
> IS_ERR_VALUE() assumes that its parameter is an unsigned long.
> It can not be used to check if an 'unsigned int' reflects an error.
> As they pass an 'unsigned int' into a function that takes an
> 'unsigned long' argument. This happens to work
From: Ivan Vecera
> Sent: 29 July 2016 18:53
> Change type of bna_id to atomic_t. The bnad_list_mutex is used to prevent
> a race when bna_id is incremented. After the change the mutex can be
> removed in the next step.
...
> -static u32 bna_id;
> +static atomic_t bna_id;
...
>
From: Baole Ni
> Sent: 02 August 2016 11:39
> I find that the developers often just specified the numeric value
> when calling a macro which is defined with a parameter for access permission.
> As we know, these numeric value for access permission have had the
> corresponding macro,
> and that
From: Phil Sutter
> Sent: 03 August 2016 22:23
> This is required to correctly interpret INET_DIAG_INFO messages exported
> by sctp_diag module.
...
> diff --git a/include/linux/sctp.h b/include/linux/sctp.h
> index de1f64318fc4e..fcb4c36461732 100644
> --- a/include/linux/sctp.h
> +++
From: Gavin Shan
> Sent: 15 July 2016 11:44
> The device is assigned with random MAC address. It isn't reasonable.
> An valid MAC address might have been provided by (uboot) firmware by
> device-tree or in chip. It's reasonable to use it to maintain consistency.
>
> This uses the MAC address from
From: Bjørn Mork
> Sent: 13 July 2016 23:23
...
> Or how about the more generic?:
>
> if (bp[0] & 0x02)
> eth_hw_addr_random(net);
> else
> ether_addr_copy(net->dev_addr, bp);
>
> That would catch similar screwups from other vendors too.
Not really,
From: Brenden Blanco
> Sent: 12 July 2016 08:51
> Add support for the BPF_PROG_TYPE_XDP hook in mlx4 driver.
>
> In tc/socket bpf programs, helpers linearize skb fragments as needed
> when the program touches the packet data. However, in the pursuit of
> speed, XDP programs will not be allowed to
From: Ivan Vecera
> Sent: 27 June 2016 10:41
> On 26.6.2016 10:59, Ben Hutchings wrote:
> > On Fri, 2016-03-18 at 13:24 +0100, Ivan Vecera wrote:
> >> Memory allocated at several places is not appropriately freed.
> >
> > Given that ethtool is not a library or a long-running application - why
> >
From: Jakub Kicinski
> Sent: 01 July 2016 22:27
>
> C bitfields are problematic and best avoided. Developers
> interacting with hardware registers find themselves searching
> for easy-to-use alternatives. Common approach is to define
> structures or sets of macros containing mask and shift
From: Vishwanath Pai
> Sent: 08 July 2016 00:34
> I am planning to add a revision 2 for the hashlimit xtables module to
> support higher packets per second rates. This patch renames all the
> functions and variables related to revision 1 by adding _v1 at the end of
> the names.
Sounds backwards.
From: Pablo Neira Ayuso
> Sent: 06 July 2016 15:24
> From: Florian Westphal
>
> increases struct size by 32 bytes (288 -> 320), but it is the right thing,
> else any attempt to (re-)arrange nf_conn members by cacheline won't work.
...
> nf_conntrack_cachep =
From: Stephen Hemminger
> Sent: 07 July 2016 05:05
> On Tue, 5 Jul 2016 10:53:38 -0500
> Eli Cohen wrote:
>
> >
> > +static int extract_guid(__u64 *guid, char *arg)
> > +{
> > + __u64 ret;
> > + int g[8];
> > + int err;
> > +
> > + err = sscanf(arg,
From: Arvind Yadav
> Sent: 07 July 2016 19:38
> -Return type of 'qe_muram_alloc' is 'unsigned long', That Was trying to
> assigned in ucc_fast_tx_virtual_fifo_base_offset and
> ucc_fast_rx_virtual_fifo_base_offset. These variable are 'unsigned int'.
> So before assginment need a proper type
From: Arvind Yadav
> Sent: 04 August 2016 17:53
> IS_ERR_VALUE() assumes that parameter is an unsigned long.
> It can not be used to check if 'unsigned int' is passed insted.
> Which tends to reflect an error.
> In 64bit architectures sizeof (int) == 4 && sizeof (long) == 8.
> IS_ERR_VALUE(x) is
From: netdev-ow...@vger.kernel.org [mailto:netdev-ow...@vger.kernel.org] On
Behalf Of Arvind Yadav
> IS_ERR_VALUE() assumes that parameter is an unsigned long.
> It can not be used to check if 'unsigned int' is passed insted.
> Which tends to reflect an error.
> In 64bit architectures sizeof
From: Arnd Bergmann
> Sent: 08 August 2016 16:13
>
> On Monday, August 8, 2016 2:49:11 PM CEST David Laight wrote:
> >
> > > If qe_muram_alloc will return any error, Then IS_ERR_VALUE will always
> > > return 0. it'll not call ucc_fast_free for any failure. Insi
From: Stephen Hemminger
> Sent: 08 August 2016 16:57
> To: Jiri Pirko
...
> > >> In order to not confuse reader, write help messages into stderr.
> > >>
> > >> Signed-off-by: Jiri Pirko
> > >
> > >This does make devlink consistent with other parts of iproute2.
> > >But the most
From: Jeff Kirsher
> Sent: 29 June 2016 05:41
> From: Mitch Williams
>
> This function uses the i40e_hw struct all over the place, so why doesn't
> it keep a pointer to the struct? Add this pointer as a local variable
> and use it consistently throughout the function.
From: Stefan Hajnoczi
> Sent: 10 August 2016 12:52
> On Mon, Aug 08, 2016 at 06:14:41PM +0200, ggar...@abra.uab.cat wrote:
> > diff --git a/include/uapi/linux/vsockmon.h b/include/uapi/linux/vsockmon.h
> > new file mode 100644
> > index 000..739b4bf
> > --- /dev/null
> > +++
From: Xin Long
> Sent: 13 August 2016 08:48
> >
> > This style of error handling is dangerous. The first error can be
> > lost.
> >
> > For example, if sctp_outq_flush_rtx() earlier in this function returns
> > an error, it will be lost if any invocation of the function
> > sctp_packet_transmit()
From: Mintz, Yuval
> Sent: 30 January 2017 17:56
> > How many different implementations of 'ops->adjfreq' are there?
> > If there is only one you don't need an indirect call.
>
> There's only one implementation. But qed publishes its functions to
> qede [and other modules] by structs of
Commit 26abe1437 changed sock_create_kern() so that it stopped
holding a reference to the network namespace.
The rational seemed to be 'to allow to stop it' (presumably 'be deleted').
Prior to this change some kernel paths used sk_change_net() (etc) to
change the namespace after the socket was
From: Saeed Mahameed
> Sent: 31 January 2017 20:59
> From: Daniel Jurgens
>
> There is a hardware feature that will pad the start or end of a DMA to
> be cache line aligned to avoid RMWs on the last cache line. The default
> cache line size setting for this feature is 64B.
From: Saeed Mahameed
> Sent: 31 January 2017 20:59
> From: Or Gerlitz
>
> For some reason, sparse doesn't like using an expression of type (!x)
> with a bitwise | and &. In order to mitigate that, we use a local
> variable.
>
> Since getting a typeof(bitfield) is
From: Saeed Mahameed
> Sent: 31 January 2017 20:59
> Cache skb_vlan_tag_present(skb) and pass it wherever needed in xmit
> routines.
...
Does this actually generate better code?
It is quite likely that your 'vlan_present' variable ends up being on stack.
Whereas the 'skb' is likely to be in a
From Florian Fainelli
> Sent: 31 January 2017 20:33
> On 01/31/2017 11:03 AM, Iyappan Subramanian wrote:
> > This patch adds,
> > - Transmit
> > - Transmit completion poll
> > - Receive poll
> > - NAPI handler
> >
> > and enables the driver.
> >
> > Signed-off-by: Iyappan
101 - 200 of 618 matches
Mail list logo