Re: [Qemu-devel] [PATCH V2 2/3] colo-compare: track connection and enqueue packet

[Zhang Chen <zhangchen.fnst@cn.fujitsu.com>]

On 03/30/2016 06:36 PM, Dr. David Alan Gilbert wrote:
> * Zhang Chen (zhangchen.fnst@cn.fujitsu.com) wrote:
>> In this patch we use kernel jhash table to track
>> connection, and then enqueue net packet like this:
>>
>> + CompareState ++
>> |               |
>> +---------------+   +---------------+         +---------------+
>> |conn list      +--->conn           +--------->conn           |
>> +---------------+   +---------------+         +---------------+
>> |               |     |           |             |          |
>> +---------------+ +---v----+  +---v----+    +---v----+ +---v----+
>>                    |primary |  |secondary    |primary | |secondary
>>                    |packet  |  |packet  +    |packet  | |packet  +
>>                    +--------+  +--------+    +--------+ +--------+
>>                        |           |             |          |
>>                    +---v----+  +---v----+    +---v----+ +---v----+
>>                    |primary |  |secondary    |primary | |secondary
>>                    |packet  |  |packet  +    |packet  | |packet  +
>>                    +--------+  +--------+    +--------+ +--------+
>>                        |           |             |          |
>>                    +---v----+  +---v----+    +---v----+ +---v----+
>>                    |primary |  |secondary    |primary | |secondary
>>                    |packet  |  |packet  +    |packet  | |packet  +
>>                    +--------+  +--------+    +--------+ +--------+
>>
>> Signed-off-by: Zhang Chen <zhangchen.fnst@cn.fujitsu.com>
>> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
>> Signed-off-by: Wen Congyang <wency@cn.fujitsu.com>
>> ---
>>   include/qemu/jhash.h |  59 ++++++++++
>>   net/colo-compare.c   | 324 ++++++++++++++++++++++++++++++++++++++++++++++++++-
>>   2 files changed, 380 insertions(+), 3 deletions(-)
>>   create mode 100644 include/qemu/jhash.h
>>
>> diff --git a/include/qemu/jhash.h b/include/qemu/jhash.h
>> new file mode 100644
>> index 0000000..8a8ff0f
>> --- /dev/null
>> +++ b/include/qemu/jhash.h
>> @@ -0,0 +1,59 @@
>> +/* jhash.h: Jenkins hash support.
>> +  *
>> +  * Copyright (C) 2006. Bob Jenkins (bob_jenkins@burtleburtle.net)
>> +  *
>> +  * http://burtleburtle.net/bob/hash/
>> +  *
>> +  * These are the credits from Bob's sources:
>> +  *
>> +  * lookup3.c, by Bob Jenkins, May 2006, Public Domain.
>> +  *
>> +  * These are functions for producing 32-bit hashes for hash table lookup.
>> +  * hashword(), hashlittle(), hashlittle2(), hashbig(), mix(), and final()
>> +  * are externally useful functions.  Routines to test the hash are included
>> +  * if SELF_TEST is defined.  You can use this free for any purpose.It's in
>> +  * the public domain.  It has no warranty.
>> +  *
>> +  * Copyright (C) 2009-2010 Jozsef Kadlecsik (kadlec@blackhole.kfki.hu)
>> +  *
>> +  * I've modified Bob's hash to be useful in the Linux kernel, and
>> +  * any bugs present are my fault.
>> +  * Jozsef
>> +  */
>> +
>> +#ifndef QEMU_JHASH_H__
>> +#define QEMU_JHASH_H__
>> +
>> +#include "qemu/bitops.h"
>> +
>> +/*
>> + * hashtable related is copied from linux kernel jhash
>> + */
>> +
>> +/* __jhash_mix -- mix 3 32-bit values reversibly. */
>> +#define __jhash_mix(a, b, c)                \
>> +{                                           \
>> +    a -= c;  a ^= rol32(c, 4);  c += b;     \
>> +    b -= a;  b ^= rol32(a, 6);  a += c;     \
>> +    c -= b;  c ^= rol32(b, 8);  b += a;     \
>> +    a -= c;  a ^= rol32(c, 16); c += b;     \
>> +    b -= a;  b ^= rol32(a, 19); a += c;     \
>> +    c -= b;  c ^= rol32(b, 4);  b += a;     \
>> +}
>> +
>> +/* __jhash_final - final mixing of 3 32-bit values (a,b,c) into c */
>> +#define __jhash_final(a, b, c)  \
>> +{                               \
>> +    c ^= b; c -= rol32(b, 14);  \
>> +    a ^= c; a -= rol32(c, 11);  \
>> +    b ^= a; b -= rol32(a, 25);  \
>> +    c ^= b; c -= rol32(b, 16);  \
>> +    a ^= c; a -= rol32(c, 4);   \
>> +    b ^= a; b -= rol32(a, 14);  \
>> +    c ^= b; c -= rol32(b, 24);  \
>> +}
>> +
>> +/* An arbitrary initial parameter */
>> +#define JHASH_INITVAL           0xdeadbeef
>> +
>> +#endif /* QEMU_JHASH_H__ */
>> diff --git a/net/colo-compare.c b/net/colo-compare.c
>> index 62c66df..0bb5a51 100644
>> --- a/net/colo-compare.c
>> +++ b/net/colo-compare.c
>> @@ -20,15 +20,22 @@
>>   #include "net/queue.h"
>>   #include "sysemu/char.h"
>>   #include "qemu/sockets.h"
>> +#include <sys/sysinfo.h>
>> +#include "slirp/slirp.h"
>> +#include "qemu/jhash.h"
>> +#include <sys/sysinfo.h>
>>   
>>   #define TYPE_COLO_COMPARE "colo-compare"
>>   #define COLO_COMPARE(obj) \
>>       OBJECT_CHECK(CompareState, (obj), TYPE_COLO_COMPARE)
>>   
>>   #define COMPARE_READ_LEN_MAX NET_BUFSIZE
>> +#define PAGE_SIZE 4096
>> +#define ETH_HLEN 14
> PAGE_SIZE is not just 4k; use one of the system headers.

OK, I will fix it with include/exec/cpu-all.h TARGET_PAGE_SIZE

> Also, don't define ETH_HLEN - include net/eth.h

OK

>>   static QTAILQ_HEAD(, CompareState) net_compares =
>>          QTAILQ_HEAD_INITIALIZER(net_compares);
>> +static ssize_t hashtable_max_size;
>>   
>>   typedef struct ReadState {
>>       int state; /* 0 = getting length, 1 = getting data */
>> @@ -37,6 +44,28 @@ typedef struct ReadState {
>>       uint8_t buf[COMPARE_READ_LEN_MAX];
>>   } ReadState;
>>   
>> +/*
>> +  + CompareState ++
>> +  |               |
>> +  +---------------+   +---------------+         +---------------+
>> +  |conn list      +--->conn           +--------->conn           |
>> +  +---------------+   +---------------+         +---------------+
>> +  |               |     |           |             |          |
>> +  +---------------+ +---v----+  +---v----+    +---v----+ +---v----+
>> +                    |primary |  |secondary    |primary | |secondary
>> +                    |packet  |  |packet  +    |packet  | |packet  +
>> +                    +--------+  +--------+    +--------+ +--------+
>> +                        |           |             |          |
>> +                    +---v----+  +---v----+    +---v----+ +---v----+
>> +                    |primary |  |secondary    |primary | |secondary
>> +                    |packet  |  |packet  +    |packet  | |packet  +
>> +                    +--------+  +--------+    +--------+ +--------+
>> +                        |           |             |          |
>> +                    +---v----+  +---v----+    +---v----+ +---v----+
>> +                    |primary |  |secondary    |primary | |secondary
>> +                    |packet  |  |packet  +    |packet  | |packet  +
>> +                    +--------+  +--------+    +--------+ +--------+
>> +*/
>>   typedef struct CompareState {
>>       Object parent;
>>   
>> @@ -49,8 +78,268 @@ typedef struct CompareState {
>>       QTAILQ_ENTRY(CompareState) next;
>>       ReadState pri_rs;
>>       ReadState sec_rs;
>> +
>> +    /* connection list: the connections belonged to this NIC could be found
>> +     * in this list.
>> +     * element type: Connection
>> +     */
>> +    GQueue conn_list;
>> +    QemuMutex conn_list_lock; /* to protect conn_list */
>> +    /* hashtable to save connection */
>> +    GHashTable *connection_track_table;
>> +    /* to save unprocessed_connections */
>> +    GQueue unprocessed_connections;
>> +    /* proxy current hash size */
>> +    ssize_t hashtable_size;
>>   } CompareState;
>>   
>> +typedef struct Packet {
>> +    void *data;
>> +    union {
>> +        uint8_t *network_layer;
>> +        struct ip *ip;
>> +    };
>> +    uint8_t *transport_layer;
>> +    int size;
>> +    CompareState *s;
>> +} Packet;
>> +
>> +typedef struct ConnectionKey {
>> +    /* (src, dst) must be grouped, in the same way than in IP header */
>> +    struct in_addr src;
>> +    struct in_addr dst;
>> +    uint16_t src_port;
>> +    uint16_t dst_port;
>> +    uint8_t ip_proto;
>> +} QEMU_PACKED ConnectionKey;
> Someone will want IPv6 at some point, so think about that, but not
> too worried for now.
>
>> +typedef struct Connection {
>> +    QemuMutex list_lock;
>> +    /* connection primary send queue: element type: Packet */
>> +    GQueue primary_list;
>> +    /* connection secondary send queue: element type: Packet */
>> +    GQueue secondary_list;
>> +    /* flag to enqueue unprocessed_connections */
>> +    bool processing;
>> +    int ip_proto;
> in ConnectionKey you use uint8_t for ip_proto  - should
> be consistent?

OK, will fix it

>
>> +} Connection;
>> +
>> +enum {
>> +    PRIMARY_IN = 0,
>> +    SECONDARY_IN,
>> +};
>> +
>> +static void packet_destroy(void *opaque, void *user_data);
>> +static int compare_chr_send(CharDriverState *out, const uint8_t *buf, int size);
>> +
>> +static uint32_t connection_key_hash(const void *opaque)
>> +{
>> +    const ConnectionKey *key = opaque;
>> +    uint32_t a, b, c;
>> +
>> +    /* Jenkins hash */
>> +    a = b = c = JHASH_INITVAL + sizeof(*key);
>> +    a += key->src.s_addr;
>> +    b += key->dst.s_addr;
>> +    c += (key->src_port | key->dst_port << 16);
>> +    __jhash_mix(a, b, c);
>> +
>> +    a += key->ip_proto;
>> +    __jhash_final(a, b, c);
>> +
>> +    return c;
>> +}
>> +
>> +static int connection_key_equal(const void *opaque1, const void *opaque2)
>> +{
>> +    return memcmp(opaque1, opaque2, sizeof(ConnectionKey)) == 0;
>> +}
>> +
>> +/*
>> + *  initialize connecon_key for packet
>                          ^ti
>
>> + *  Return 0 on success, if return 1 the pkt will be sent later
>> + */
>> +static int connection_key_init(Packet *pkt, ConnectionKey *key)
>> +{
>> +    int network_length;
>> +    uint8_t *data = pkt->data;
>> +    uint16_t l3_proto;
>> +    uint32_t tmp_ports;
>> +    ssize_t l2hdr_len = eth_get_l2_hdr_length(data);
>> +
>> +    pkt->network_layer = data + ETH_HLEN;
>> +    l3_proto = eth_get_l3_proto(data, l2hdr_len);
>> +    if (l3_proto != ETH_P_IP) {
>> +        return 1;
>> +    }
>> +
>> +    network_length = pkt->ip->ip_hl * 4;
>> +    pkt->transport_layer = pkt->network_layer + network_length;
> Have we checked that this is valid - this is guest/external network
> data, so is that 'network_length' actually pointing to valid data
> or off the end of the packet?

I will check it before use in next version.

>
>> +    key->ip_proto = pkt->ip->ip_p;
>> +    key->src = pkt->ip->ip_src;
>> +    key->dst = pkt->ip->ip_dst;
>> +
>> +    switch (key->ip_proto) {
>> +    case IPPROTO_TCP:
>> +    case IPPROTO_UDP:
>> +    case IPPROTO_DCCP:
>> +    case IPPROTO_ESP:
>> +    case IPPROTO_SCTP:
>> +    case IPPROTO_UDPLITE:
>> +        tmp_ports = *(uint32_t *)(pkt->transport_layer);
>> +        key->src_port = tmp_ports & 0xffff;
>> +        key->dst_port = tmp_ports >> 16;
> Do these need ntohs - or do you want to keep them in network
> order?  In my world on your older code I added ntohs's because
> it made debugging make a lot more sense when you print out src_port/dst_port.

Make sense, I will fix it.

>
>> +        break;
>> +    case IPPROTO_AH:
>> +        tmp_ports = *(uint32_t *)(pkt->transport_layer + 4);
>> +        key->src_port = tmp_ports & 0xffff;
>> +        key->dst_port = tmp_ports >> 16;
>> +        break;
>> +    default:
>     Do you need to set src_port/dst_port here (to 0 ?? ) ?

Yes,will fix it.

>
>> +        break;
>> +    }
>> +
>> +    return 0;
>> +}
>> +
>> +static Connection *connection_new(ConnectionKey *key)
>> +{
>> +    Connection *conn = g_slice_new(Connection);
>> +
>> +    qemu_mutex_init(&conn->list_lock);
>> +    conn->ip_proto = key->ip_proto;
>> +    conn->processing = false;
>> +    g_queue_init(&conn->primary_list);
>> +    g_queue_init(&conn->secondary_list);
>> +
>> +    return conn;
>> +}
>> +
>> +/*
>> + * Clear hashtable, stop this hash growing really huge
>> + */
>> +static void connection_hashtable_reset(CompareState *s)
>> +{
>> +    s->hashtable_size = 0;
>> +    g_hash_table_remove_all(s->connection_track_table);
>> +}
>> +
>> +/* if not found, creata a new connection and add to hash table */
>   Typo                    ^

will fix.

>> +static Connection *connection_get(CompareState *s, ConnectionKey *key)
>> +{
>> +    /* FIXME: protect connection_track_table */
>> +    Connection *conn = g_hash_table_lookup(s->connection_track_table, key);
>> +
>> +    if (conn == NULL) {
>> +        ConnectionKey *new_key = g_memdup(key, sizeof(*key));
>> +
>> +        conn = connection_new(key);
>> +
>> +        s->hashtable_size++;
>> +        if (s->hashtable_size > hashtable_max_size) {
>> +            error_report("colo proxy connection hashtable full, clear it");
>> +            connection_hashtable_reset(s);
>> +            /* TODO:clear conn_list */
>> +        } else {
> This feels wrong; should this actually be in an else? If you've just cleared
> the hash table, then you probably want to add this new connection to the empty
> table? (And for example at the moment the 'new_key' is not used if we go down
> this if).
>

you are right, I will fix.

>> +            g_hash_table_insert(s->connection_track_table, new_key, conn);
>> +        }
>> +    }
>> +
>> +     return conn;
>> +}
>> +
>> +static void connection_destroy(void *opaque)
>> +{
>> +    Connection *conn = opaque;
>> +
>> +    qemu_mutex_lock(&conn->list_lock);
>> +    g_queue_foreach(&conn->primary_list, packet_destroy, NULL);
>> +    g_queue_free(&conn->primary_list);
>> +    g_queue_foreach(&conn->secondary_list, packet_destroy, NULL);
>> +    g_queue_free(&conn->secondary_list);
>> +    qemu_mutex_unlock(&conn->list_lock);
>> +    qemu_mutex_destroy(&conn->list_lock);
>> +    g_slice_free(Connection, conn);
>> +}
>> +
>> +static Packet *packet_new(CompareState *s, const void *data,
>> +                              int size, ConnectionKey *key)
>> +{
>> +    Packet *pkt = g_slice_new(Packet);
>> +
>> +    pkt->data = g_memdup(data, size);
>> +    pkt->size = size;
>> +    pkt->s = s;
>> +
>> +    if (connection_key_init(pkt, key)) {
>> +        packet_destroy(pkt, NULL);
>> +        pkt = NULL;
>> +    }
>> +
>> +    return pkt;
>> +}
>> +
>> +static int packet_enqueue(CompareState *s, int mode)
>> +{
>> +    ConnectionKey key = {{ 0 } };
>> +    Packet *pkt = NULL;
>> +    Connection *conn;
>> +
>> +    /* arp packet will be sent */
> Can you add some more detail about that - what do the return
> values of packet_enqueue mean; what happens to things like IPv6 or ARP packets?

OK, will add some comments.
If the primary input packet is not IP packet, we will send it for now.


>
>> +    if (mode == PRIMARY_IN) {
>> +        pkt = packet_new(s, s->pri_rs.buf, s->pri_rs.packet_len, &key);
>> +    } else {
>> +        pkt = packet_new(s, s->sec_rs.buf, s->sec_rs.packet_len, &key);
>> +    }
>> +    if (!pkt) {
>> +        return -1;
>> +    }
>> +
>> +    conn = connection_get(s, &key);
>> +    if (!conn->processing) {
>> +        qemu_mutex_lock(&s->conn_list_lock);
>> +        g_queue_push_tail(&s->conn_list, conn);
>> +        qemu_mutex_unlock(&s->conn_list_lock);
>> +        conn->processing = true;
>> +    }
>> +
>> +    qemu_mutex_lock(&conn->list_lock);
>> +    if (mode == PRIMARY_IN) {
>> +        g_queue_push_tail(&conn->primary_list, pkt);
>> +    } else {
>> +        g_queue_push_tail(&conn->secondary_list, pkt);
>> +    }
>> +    qemu_mutex_unlock(&conn->list_lock);
>> +
>> +    return 0;
>> +}
>> +
>> +static void packet_destroy(void *opaque, void *user_data)
>> +{
>> +    Packet *pkt = opaque;
>> +
>> +    g_free(pkt->data);
>> +    g_slice_free(Packet, pkt);
>> +}
>> +
>> +static inline void colo_flush_connection(void *opaque, void *user_data)
>> +{
> Is this used?

will be used, like lizhijian said.

>
>> +    Connection *conn = opaque;
>> +    Packet *pkt = NULL;
>> +
>> +    qemu_mutex_lock(&conn->list_lock);
>> +    while (!g_queue_is_empty(&conn->primary_list)) {
>> +        pkt = g_queue_pop_head(&conn->primary_list);
>> +        compare_chr_send(pkt->s->chr_out, pkt->data, pkt->size);
>> +        /* FIXME: destroy pkt ?*/
>> +    }
>> +    while (!g_queue_is_empty(&conn->secondary_list)) {
>> +        pkt = g_queue_pop_head(&conn->secondary_list);
>> +        packet_destroy(pkt, NULL);
>> +    }
>> +    qemu_mutex_unlock(&conn->list_lock);
>> +}
>> +
>>   static int compare_chr_send(CharDriverState *out, const uint8_t *buf, int size)
>>   {
>>       int ret = 0;
>> @@ -142,8 +431,10 @@ static void compare_pri_chr_in(void *opaque, const uint8_t *buf, int size)
>>   
>>       ret = compare_chr_fill_rstate(&s->pri_rs, buf, size);
>>       if (ret == 1) {
>> -        /* FIXME: enqueue to primary packet list */
>> -        compare_chr_send(s->chr_out, buf, size);
>> +        if (packet_enqueue(s, PRIMARY_IN)) {
>> +            error_report("primary: unsupported packet in");
> Is this for non-IP packets?  If so you don't want an error_report - because non-IP are
> quite common; a trace would be useful giving the packet type etc
>
>> +            compare_chr_send(s->chr_out, buf, size);
>> +        }
>>       } else if (ret == -1) {
>>           qemu_chr_add_handlers(s->chr_pri_in, NULL, NULL, NULL, NULL);
>>       }
>> @@ -156,7 +447,9 @@ static void compare_sec_chr_in(void *opaque, const uint8_t *buf, int size)
>>   
>>       ret = compare_chr_fill_rstate(&s->sec_rs, buf, size);
>>       if (ret == 1) {
>> -        /* TODO: enqueue to secondary packet list*/
>> +        if (packet_enqueue(s, SECONDARY_IN)) {
>> +            error_report("secondary: unsupported packet in");
>> +        }
>>       } else if (ret == -1) {
>>           qemu_chr_add_handlers(s->chr_sec_in, NULL, NULL, NULL, NULL);
>>       }
>> @@ -210,6 +503,7 @@ static void compare_set_outdev(Object *obj, const char *value, Error **errp)
>>   static void colo_compare_complete(UserCreatable *uc, Error **errp)
>>   {
>>       CompareState *s = COLO_COMPARE(uc);
>> +    struct sysinfo si;
>>   
>>       if (!s->pri_indev || !s->sec_indev || !s->outdev) {
>>           error_setg(errp, "colo compare needs 'primary_in' ,"
>> @@ -255,6 +549,29 @@ static void colo_compare_complete(UserCreatable *uc, Error **errp)
>>   
>>       QTAILQ_INSERT_TAIL(&net_compares, s, next);
>>   
>> +    g_queue_init(&s->conn_list);
>> +    qemu_mutex_init(&s->conn_list_lock);
>> +
>> +    s->hashtable_size = 0;
>> +    /*
>> +     * Idea from kernel tcp.c: use 1/16384 of memory.  On i386: 32MB
>> +     * machine has 512 buckets. >= 1GB machines have 16384 buckets.
>> +     */
>> +    sysinfo(&si);
>> +    hashtable_max_size = si.totalram / 16384;
>> +    if (si.totalram > (1024 * 1024 * 1024 / PAGE_SIZE)) {
>> +        hashtable_max_size = 16384;
>> +    }
>> +    if (hashtable_max_size < 32) {
>> +        hashtable_max_size = 32;
>> +    }
>> +    hashtable_max_size = hashtable_max_size * 8; /* default factor = 8 */
> Make this a lot simpler; just pick a size and if it's a problem then we'll worry
> about it later, or make it an option on the filter if you want it changeable.
>

OK, I will add hashtable_max_size to make it changeable.

Thanks
zhangchen

>> +    s->connection_track_table = g_hash_table_new_full(connection_key_hash,
>> +                                                      connection_key_equal,
>> +                                                      g_free,
>> +                                                      connection_destroy);
>> +
>>       return;
>>   
>>   out:
>> @@ -297,6 +614,7 @@ static void colo_compare_class_finalize(ObjectClass *oc, void *data)
>>       if (!QTAILQ_EMPTY(&net_compares)) {
>>           QTAILQ_REMOVE(&net_compares, s, next);
>>       }
>> +    qemu_mutex_destroy(&s->conn_list_lock);
>>   }
>>   
>>   static void colo_compare_init(Object *obj)
>> -- 
>> 1.9.1
>>
>>
>>
> --
> Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
>
>
> .
>

-- 
Thanks
zhangchen