a crash on debug image (VPP-1151)
In debug image, there is extra code to validate the buffer when it is
freed. It uses the hash table to lookup the buffer index with spinlock
to prevent contention. However, there is one spinlock for each worker
thread. So allocating the buffer on thread-x and freeing the same buffer
on thread-y causes the validation to fail on thread-y. The fix is to
have only one spinlock, stored in vlib_global_main.
Change-Id: Ic383846cefe84a3e262255afcf82276742f0f62e
Signed-off-by: Steven <sluong@cisco.com>
Also fixes old ixge driver, so it works with recent physmem changes and
vfio.
Change-Id: Id4be74b34daed47cd281a77eec43d6692340d882
Signed-off-by: Damjan Marion <damarion@cisco.com>
Currently, every alloc involves callback function call. With this change
callback function is called only if there is no empty buffers on the
free list.
Change-Id: I2238c19ece7ce182c49ba0f2485add52f335f3b6
Signed-off-by: Damjan Marion <damarion@cisco.com>
This change avoids hitting
ASSERT (vec_len (f->buffers) == f->n_alloc);
in vlib_buffer_delete_free_list_internal by updating
vlib_buffer_free_list_t->n_alloc in case when buffers are removed from
the free list due to overfill.
Change-Id: Id60fa97e8cb0f02ef529b05816265a57a4e7d998
Signed-off-by: Klement Sekera <ksekera@cisco.com>
This patch adds supprot support for multiple numa-aware physmem regions.
Change-Id: I5c69a6f4da33c8ee21bdb8604d52fd2886f2327e
Signed-off-by: Damjan Marion <damarion@cisco.com>
First attempt to make internal buffer manager thread safe was not
succesfull, so trying again. This time with more testing.
Change-Id: I01b8385a9c26d233934a3339255ea4bd31c865ac
Signed-off-by: Damjan Marion <damarion@cisco.com>
To save space in the first cacheline following is changed:
- total_length_not_including_first_buffer moved to the 2nd cacheline.
This field is used only when VLIB_BUFFER_TOTAL_LENGTH_VALID and
VLIB_BUFFER_NEXT_PRESENT are both set.
- free_list_index is now stored in 4bits inside flags, which
allows up to 16 free lists. In case we need more we can store index
in the 2nd cachelin
Change-Id: Ic8521350819391af470d31d3fa1013e67ecb7681
Signed-off-by: Damjan Marion <damarion@cisco.com>
- vnet_buffer_pool_create should probably be named
dpdk_buffer_pool_create since that is what it does.
- Its prototype should also be in a DPDK plugin header, not in
vlib/buffer_funcs.h, since the implementation is in the plugin
and nobody else should be calling it.
Change-Id: I7ba259afa4b888bc94f3ad257305e286b41e7370
Signed-off-by: Chris Luke <chrisy@flirble.org>
Currently, buffer index is calculated as a offset to the physmem
region shifted by log2_cacheline size.
When DPDK is used we "hack" physmem data with information taken from
dpdk mempool. This makes physmem code not usable with DPDK.
This change makes buffer memory start and size independent of physmem
basically allowing physmem to be used when DPDK plugin is loaded.
Change-Id: Ieb399d398f147583b9baab467152a352d58c9c31
Signed-off-by: Damjan Marion <damarion@cisco.com>
This patch deprecates stack-based thread identification,
Also removes requirement that thread stacks are adjacent.
Finally, possibly annoying for some folks, it renames
all occurences of cpu_index and cpu_number with thread
index. Using word "cpu" is misleading here as thread can
be migrated ti different CPU, and also it is not related
to linux cpu index.
Change-Id: I68cdaf661e701d2336fc953dcb9978d10a70f7c1
Signed-off-by: Damjan Marion <damarion@cisco.com>
