At 2020-08-03 20:34:25, "Olivier Matz" [off-list ref] wrote:

On Mon, Aug 03, 2020 at 05:42:13PM +0800, yang_y_yi wrote:

quoted

At 2020-08-03 16:11:39, "Olivier Matz" [off-list ref] wrote:

quoted

On Mon, Aug 03, 2020 at 09:26:40AM +0800, yang_y_yi wrote:

quoted

At 2020-08-03 04:29:07, "Olivier Matz" [off-list ref] wrote:

quoted

Hi,

On Sun, Aug 02, 2020 at 07:12:36AM +0800, yang_y_yi wrote:

quoted

At 2020-07-31 23:15:43, "Olivier Matz" [off-list ref] wrote:

quoted

Hi,

On Thu, Jul 30, 2020 at 08:08:59PM +0800, yang_y_yi@163.com wrote:

quoted

From: Yi Yang <redacted>

In GSO case, segmented mbufs are attached to original
mbuf which can't be freed when it is external. The issue
is free_cb doesn't know original mbuf and doesn't free
it when refcnt of shinfo is 0.

Original mbuf can be freed by rte_pktmbuf_free segmented
mbufs or by rte_pktmbuf_free original mbuf. Two kind of
cases should have different behaviors. free_cb won't
explicitly call rte_pktmbuf_free to free original mbuf
if it is freed by rte_pktmbuf_free original mbuf, but it
has to call rte_pktmbuf_free to free original mbuf if it
is freed by rte_pktmbuf_free segmented mbufs.

In order to fix this issue, free_cb interface has to been
changed, __rte_pktmbuf_free_extbuf must deliver called
mbuf pointer to free_cb, argument opaque can be defined
as a custom struct by user, it can includes original mbuf
pointer, user-defined free_cb can compare caller mbuf with
mbuf in opaque struct, free_cb should free original mbuf
if they are not same, this corresponds to rte_pktmbuf_free
segmented mbufs case, otherwise, free_cb won't free original
mbuf because the caller explicitly called rte_pktmbuf_free
to free it.

Here is pseduo code to show two kind of cases.

case 1. rte_pktmbuf_free segmented mbufs

nb_tx = rte_gso_segment(original_mbuf, /* original mbuf */
                      &gso_ctx,
                      /* segmented mbuf */
                      (struct rte_mbuf **)&gso_mbufs,
                      MAX_GSO_MBUFS);

I'm sorry but it is not very clear to me what operations are done by
rte_gso_segment().

In the current code, I only see calls to rte_pktmbuf_attach(),
which do not deal with external buffers. Am I missing something?

Are you able to show the issue only with mbuf functions? It would
be helpful to understand what does not work.


Thanks,
Olivier

Oliver, thank you for comment, let me show you why it doesn't work for my use case.  In OVS DPDK, VM uses vhostuserclient to send large packets whose size is about 64K because we enabled TSO & UFO, these large packets use rte_mbufs allocated by DPDK virtio_net function 
virtio_dev_pktmbuf_alloc() (in file lib/librte_vhost/virtio_net.c. Please refer to [PATCH V1 3/3], I changed free_cb as below, these packets use the same allocate function and the same free_cb no matter they are TCP packet or UDP packets, in case of VXLAN TSO, most NICs can't support inner UDP fragment offload, so OVS DPDK has to do it by software, for UDP case, the original rte_mbuf only can be freed by segmented rte_mbufs which are output packets of rte_gso_segment, i.e. the original rte_mbuf only can freed by free_cb, you can see, it explicitly called rte_pktmbuf_free(arg->mbuf), the condition statement "if (caller_m != arg->mbuf)" is true for this case, this has no problem, but for TCP case, the original mbuf is delivered to rte_eth_tx_burst() but not segmented rte_mbufs output by rte_gso_segment, PMD driver will call rte_pktmbuf_free(original_rte_mbuf) but not rte_pktmbuf_free(segmented_rte_mbufs), the same free_cb will be called, that means original_rte_mbuf will be freed twice, you know what will happen, this is just the issue I'm fixing. I bring in caller_m argument, it can help work around this because caller_m is arg->mbuf and the condition statement "if (caller_m != arg->mbuf)" is false, you can't fix it without the change this patch series did.

I'm sill not sure to get your issue. Please, if you have a simple test
case using only mbufs functions (without virtio, gso, ...), it would be
very helpful because we will be sure that we are talking about the same
thing. In case there is an issue, it can easily become a unit test.

Oliver, I think you don't get the point, free operation can't be controlled by the application itself, 
it is done by PMD driver and triggered by rte_eth_tx_burst, I have shown pseudo code,
rte_gso_segment just segments a large mbuf to multiple mbufs, it won't send them, the application
will call rte_eth_tx_burst to send them finally.

quoted

That said, I looked at vhost mbuf allocation and gso segmentation, and
I found some strange things:

1/ In virtio_dev_extbuf_alloc(), and I there are 2 paths to create the
  ext mbuf.

  a/ The first one stores the shinfo struct in the mbuf, basically
     like this:

pkt = rte_pktmbuf_alloc(mp);
shinfo = rte_pktmbuf_mtod(pkt, struct rte_mbuf_ext_shared_info *);
buf = rte_malloc(NULL, buf_len, RTE_CACHE_LINE_SIZE);
shinfo->free_cb = virtio_dev_extbuf_free;
shinfo->fcb_opaque = buf;
rte_mbuf_ext_refcnt_set(shinfo, 1);

     I don't think it is a good idea, because there is no guarantee that
     the mbuf won't be freed before the buffer. For instance, doing
     this will probably fail:

pkt2 = rte_pktmbuf_alloc(mp);
rte_pktmbuf_attach(pkt2, pkt);
rte_pktmbuf_free(pkt);  /* pkt is freed, but it contains shinfo ! */

pkt is created by the application I can control, so I can control it where it will be freed, right?

This example shows that mbufs allocated like this by the vhost
driver are not constructed correctly. If an application attach a new
packet (pkt2) to it and frees the original one (pkt), it may result in a
memory corruption.

Of course, to be tested and confirmed.

No, attach will increase refcnt of shinfo, free_cb only is called when  refcnt of shinfo is decreased to
0, isn't it?

When pkt will be freed, it will decrement the shinfo refcnt, and
after it will be 1. So the buffer won't be freed. After that, the
mbuf pkt will be detached, and will return to the mbuf pool. It means
it can be reallocated, and the next user can overwrite shinfo which
is still stored in the mbuf data.

I think this is an issue of DPDK itself, if external buffer in shinfo is freed, shinfo should be set to NULL, if user will
overwrite it, he/she should use the same way as a new external buffer is attached. 

I did a test to show it, see:
http://git.droids-corp.org/?p=dpdk.git;a=commitdiff;h=a617494eeb01ff

If you run the mbuf autotest, it segfaults.

I think your test is wrong, you're changing shinfo (which is being used) in wrong way, if free_bc is NULL, it will be invalid.

static inline void
rte_pktmbuf_attach_extbuf(struct rte_mbuf *m, void *buf_addr,
        rte_iova_t buf_iova, uint16_t buf_len,
        struct rte_mbuf_ext_shared_info *shinfo)
{
        /* mbuf should not be read-only */
        RTE_ASSERT(RTE_MBUF_DIRECT(m) && rte_mbuf_refcnt_read(m) == 1);
        RTE_ASSERT(shinfo->free_cb != NULL);

For any shinfo operation, you should do it by rte_pktmbuf_attach_extbuf, you can't change it at will after that.

quoted

quoted

quoted

quoted

     To do this properly, the mbuf refcnt should be increased, and
     the mbuf should be freed in the callback. But I don't think it's
     worth doing it, given the second path (b/) looks good to me.

  b/ The second path stores the shinfo struct at the end of the
     allocated buffer, like this:

pkt = rte_pktmbuf_alloc(mp);
buf_len += sizeof(*shinfo) + sizeof(uintptr_t);
buf_len = RTE_ALIGN_CEIL(total_len, sizeof(uintptr_t));
buf = rte_malloc(NULL, buf_len, RTE_CACHE_LINE_SIZE);
shinfo = rte_pktmbuf_ext_shinfo_init_helper(buf, &buf_len,
				      virtio_dev_extbuf_free, buf);

     I think this is correct, because we have the guarantee that shinfo
     exists as long as the buffer exists.

What buffer does the allocated buffer you're saying here? The issue we're discussing how we can
free original mbuf which owns shinfo buffer.

I don't get your question.

I'm just saying that this code path looks correct, compared to
the previous one.

I think you're challenging principle of external mbuf, that isn't the thing I address.

I'm not challenging anything, I'm saying there is a bug in this code,
and the unit test above tends to confirm it.

If it is bug, you can post a patch to fix it,  it isn't related with my patches. But in my opinion, you don't
use it in correct way, I don't think it is a bug.

quoted

quoted

quoted

quoted

2/ in rte_gso_segment(), there is a loop like this:

while (pkt_seg) {
	rte_mbuf_refcnt_update(pkt_seg, -1);
	pkt_seg = pkt_seg->next;
}

  You change it to take in account the refcnt for ext mbufs.

  I may have missed something but I wonder why it is not simply:

rte_pktmbuf_free(pkt_seg);

  It will decrease the proper refcnt, and free the mbufs if they
  are not used.

Again, rte_gso_segment just decreases refcnt by one, this will ensure the last segmented 
mbuf free will trigger freeing original mbuf (only free_cb can do this).

rte_pktmbuf_free() will also decerase the refcnt, and free the resources
when the refcnt reaches 0.

It has some advantages compared to decrease the reference counter of all
segments:

- no need to iterate the segments, there is only one function call
- no need to have a special case for ext mbufs like you did in your patch
- it may be safer, in case some segments have a refcnt == 1, because
 resources will be freed.

For external mbuf, attach only increases refcnt of shinfo, refcnt of mbuf won't be touched. For normal
mbuf, attach only increase refcnt of mbuf, no shinfo there, no refcnt of shinfo increased.

I suppose rte_gso_segment() can take any mbuf type as input: standard
mbuf, indirect mbuf, ext mbuf, or even a mbuf chaing containing segments of
different types.

For instance, if you pass a chain of 2 mbufs:
- the first one is a direct mbuf containing the IP/TCP headers (orig_hdr)
- the second on is a mbuf pointing to an ext buffer (orig_payload)

I expect that the resulting mbuf after calling gso contains a list of mbufs
like this:
- a first segment containing the IP/TCP headers (new_hdr)
- a payload segment pointing on the same ext buffer

In theory, there is no reason that orig_hdr should be referenced by
another new mbuf, because it only contains headers (no data). If that's
the case, its refcnt is 1, and decreasing it to 0 without freeing it
is a bug.

For this user scenario, orig_m is owner of external buffer, small segmented mbufs reference
it, you shouldn't free orig_m before all attached segmented mbufs are freed, isn't it?

Anyway, there is maybe no issue in that case, but I was just suggesting
that using rte_pktmbuf_free() is easier to read, and safer than manually
decreasing the refcnt of each segment.

quoted

quoted

quoted

quoted

Again, sorry if this is not the issue your are referring to, but
in this case I really think that having a simple example code that
shows the issue would help.

Oliver, my statement in the patch I sent out has pseudo code to show this.  I don't think a simple
unit test can show it.

I don't see why. The PMDs and the libraries use the mbuf functions, why
a unit test couldn't call the same functions?

quoted

Let me summarize it here again. For original mbuf, there are two cases freeing
it, case one is PMD driver calls free against segmented mbufs, last segmented mbuf free will trigger
free_cb call which will free original large & extended mbuf.

OK

quoted

Case two is PMD driver will call free against
original mbuf, that also will call free_cb to free attached extended buffer.

OK

And what makes that case 1 or case 2 is executed?

quoted

In case one free_cb must call
rte_pktmbuf_free otherwise nobody will free original large & extended mbuf, in case two free_cb can't 
call rte_pktmbuf_free because the caller calling it is just rte_pktmbuf_free we need. That is to say, you
must use the same free_cb to handle these two cases, this is my issue and the point you don't get.

I think there is no need to change the free_cb API. It should work like
this:

- virtio creates the original external mbuf (orig_m)
- gso will create a new mbuf referencing the external buffer (new_m)

At this point, the shinfo has a refcnt of 2. The large buffer will be
freed as soon as rte_pktmbuf_free() is called on orig_m and new_m,
whatever the order.

Regards,
Olivier

Oliver, the reason it works is I changed free_cb API, case 1 doesn't know orig_m, how you make it free orig_m in free_cb.
The intention I change free_cb is to let it know orig_m, I saw OVS DPDK ran out out buffers and orig_m isn't freed, that is why
I want to bring in this to fix the issue. Again, in case 1, nobody explicitly calls ret_pktmbuf_free(orig_m) except free_cb I changed.

If nobody calls ret_pktmbuf_free(orig_m), it is a problem.
The free_cb is to free the buffer, not the mbuf.

To me, it should work like this:

1- virtio creates a mbuf attached to the ext buffer (the shinfo placement
  bug should be fixed)
2- gso create mbufs that reference the the same ext buf (by attaching the
  new mbuf)
3- gso must free the original mbuf

This is impossible, segmented mbufs are referencing external buffer in original mbuf,
how do you free it? As I said rte_gso_segment has no way to free it, please tell me a way if
you know how to do this.

4- the PMD transmits the new mbufs, and frees them

Whatever 3- or 4- is executed first, at the end we are sure that:
- all mbufs will be returned to the pool
- the linear buffer will be freed when the refcnt reaches 0.

If this is still unclear, please, write a unit test like I did
above to show your issue.

Regards,
Olivier

The issue is in "3- gso must free the original mbuf", rte_pktmbuf_free(segmented_mbus) can't do it,
rte_gso_segment is impossible to do it, only feasible point is free_cb, please let me know if you have
a better way to free original mbuf and don't impact on segmented mbufs in PMD. My point is you must
have a place to call rte_pktmbuf_free(rogin_m) explicitly, otherwise it is impossible  to return it to memory
pool, please point out  where it can be called in my user scenario. I don't care how it is done, I just care it can
fix my issue, please don't hesitate and send me a patch if you can, thanks a lot.

quoted

free_cb must handle case 1 and case 2 in the same code, for case 1, caller_m is segmented new_m, for case 2, caller_m is orig_m.

loop in rte_gso_segement is handling original mbuf (this mbuf is multi-mbuf and includes multiple mbufs which are linked by next
pointer), it isn't a problem at all.

Please show me code how you can fix my issue if you don't change free_cb, thank you.

struct shinfo_arg {
       void *buf;
       struct rte_mbuf *mbuf;
};

virtio_dev_extbuf_free(struct rte_mbuf *caller_m, void *opaque)
{
       struct shinfo_arg *arg = (struct shinfo_arg *)opaque;

       rte_free(arg->buf);
       if (caller_m != arg->mbuf)
               rte_pktmbuf_free(arg->mbuf);
       rte_free(arg);
}