| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Since libnet is now linked to Paflof directly, we do not have to
link it into net-snk anymore. So for board-qemu, we can now even
exclude net-snk completely from the build (for board-js2x, it is
still required for the biosemu, so we can not erase the net-snk
folder completely yet).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Now that we link libnet to Paflof, we can call the ping
function there directly instead of using the one from
net-snk. We add a similar Forth-to-C wrapper like it has
already been done for netboot() - simplification and clean-up
will be done in a later patch once we do not link against
net-snk anymore.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Now that all necessary functions are provided by Paflof, too,
we can finally link the libnet code to this binary. To be able
to call the netboot() function from the Forth code now, we also
add a wrapper that takes the parameter string from the obp-tftp
package and converts it to an argv array that is expected by
the netboot() function.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The recv() and send() functions are currently assuming that the
dictionary of the NIC node is available, so that they can call
"read" or "write" directly. While this is working fine when only
doing network loading from the Open Firmware prompt, this does
not work anymore when a client like yaboot is trying to load
additional files.
Fix this issue by looking up the execution tokens of "read" and
"write" while opening the socket, so we can use these XT values
during recv() and send() later.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Add support for virtio serial device to be used as a console device.
Currently, SLOF only supports spapr-vty device. With this addition
virtio console can be used during boot.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Initial display output is replayed to vga/vnc display, similar facility
will be needed for virtio-serial console as the enumeration would only
happen during PCI discovery.
Generalize the write routine that could be used accordingly for
virtio-serial driver as well.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Make uniform indentation(width = 4) everywhere in the file.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
For disks bigger than 2TB(512B sector size), read-10 would fail as it is
limited by the block address(4bytes). Add and use SCSI command READ(16)
when max blocks is more than 0xFFFFFFFF.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
As soon as we are booting with at least one PCI device, the
rtas-do-config-@ Forth word is called quite often, so it
makes sense that we look up the corresponding RTAS token
only once instead of each time the function is called.
Also the "ffffffff and" operation on the lower half of the
PUID is not really required and can be removed, since the
"l!" Forth word is storing the lower four bytes only anyway.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Unfortunately netapps.h currently can't be included from .code
Forth-to-C wrapper files - it includes tftp.h which in turn
includes ipv6.h, and that header contains some constructs which
can't be used in .code files. So let's make netapps.h independent
from tftp.h and add some "extern" keywords like it is done with
the other header files already that are included from .code files.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
These functions do not exist anymore, thus their prototypes
can be removed nowadays.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Commit 7e31382cca5fb988b ("Improve printed text when booting
via network") moved the "Requesting information via DHCP..."
printf statement from the netboot() function into the dhcp()
function. However, it did not take into account that the
ping() function should be changed in the same way, so with
ping, the message is currently printed out twice. So let's
re-arrange the printout of ping() now accordingly.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
If libnet is linked to the net-snk, it does not matter since the
whole stack is completely reloaded each time. But if we are
linking libnet to Paflof, we've got to make sure to properly
release the resources that we've allocated before, since the
code and data stays in memory.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
For disks bigger than 2TB(512B sector size), read-capacity-10 would fail
and return FFFFFFFF, as it only has 4byte block address field. Detect
this and use read-capacity-16 which returns 8byte block address.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
SLOF currently fails to correctly initialize the secondary and
subordinate bus number registers in the config space of PCI
bridges, so that for example with the following command line,
none of the PCI devices is usable:
qemu-system-ppc64 -nodefaults -nographic -serial mon:stdio \
-device pci-bridge,chassis_nr=1,id=bridge0,addr=0x3 \
-device pci-bridge,chassis_nr=2,id=bridge1,addr=0x4 \
-device virtio-balloon,bus=bridge1,addr=0x1 \
-device virtio-net,bus=bridge0,addr=0x2 \
-device virtio-rng,bus=bridge0,addr=0x5 \
-device pci-bridge,chassis_nr=3,id=br2,addr=0x6,bus=bridge1 \
-device e1000,bus=br2,addr=0x1
This is because SLOF tries to enumerate the PCI bus numbers
that are reachable via a bridge. In the function pci-bridge-probe,
it increases the pci-bus-number counter and writes that value into
the secondary bus number register of the PCI config space, and
after probing all attached bridges, it fills the number of the
last enumerated bus number into the subordinate bus number register.
This works fine if the whole bus enumeration is done by SLOF,
however on board-qemu, we nowadays rely on the pre-initialized PCI
device tree from QEMU - and the numbers that SLOF is trying to use
here do not match with the bus numbers that QEMU already assigned
to the bus segments (QEMU provides the device tree nodes in
descending order, but SLOF tries to enumerate the bus numbers in
ascending order here instead).
To fix this issue, we should simply stop setting up the secondary
and subordinate config space registers of the bridge in SLOF - since
this is done by QEMU already! Thus we replace the "pci-bridge-probe"
function with a board-qemu-specific function "phb-pci-bridge-probe",
that does not call "pci-bus!" and "pci-bus-subo!" anymore. (And since
pci-bridge-probe was the only spot that called phb-pci-probe-bus, we
can get rid of that wrapper now, too, and call phb-pci-walk-bridge
from phb-pci-bridge-probe directly instead).
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1377083
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The sequence "my-space pci-htype@ pci-out" in phb-pci-walk-bridge
is bugged: pci-htype@ already consumes the my-space item from the
stack, only leaving one item for pci-out. But pci-out needs two
input items on the stack, the PCI address and a character item.
So this rather should be "my-space dup pci-htype@ pci-out" instead.
However, using the output of pci-htype@ as input character for
pci-out also does not make much sense, since this is likely an
unprintable character. So let's simply use a question mark here
instead to indicate that we did not recognize the type of the
PCI device.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1377083
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The libnet code uses read() to check for keyboard input
(so that it can abort the network loading when the user
pressed ESC). So to be able to use the libnet code with
Paflof, too, we have got to provide a read() function here.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Code is slightly based on the implementation from net-snk, but
has been adapted to use the forth_eval() and forth_push/pop() macros
instead.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
They are needed for libnet, too. This implementation uses
SLOF_GetTimer() instead of using the decrementer like it is
done in the net-snk functions.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The code in lib/libnet/netload.c uses write_mm_log() to write
error messages to the management module log, thus we need to
provide this function in Paflof, too, when we want to link
the netload code to Paflof later.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
sbrk() is needed for the malloc() implementation of our libc,
so to be able to use malloc() from the Paflof code, we need
to provide sbrk() here, too.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
It's cumbersome to change the CFLAGS in multiple places in case they
have to be changed. So let's use the global CFLAGS from make.rules
for building Paflof, too. Since the global rules use some additional
compiler warning flags, fix the now occuring compiler warnings in
the Paflof code, too (mostly about missing or wrong prototypes).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Commit 2fed5652819ad26627a8 ("Always include evaluator, move
framebuffer token init to fbuffer.fs") made sure that the FCode
evaluator is always included, during each boot cycle. The basic
idea was that we would soon be starting to support PCI cards with
FCode drivers on them. However, this has never happened, and so
this change was in vain. The bad thing is now that the inclusion
of the FCode evaluator also takes a lot of precious boot time,
e.g. when running in QEMU TCG mode, it is more than a second.
So to be able to boot faster again, disable the FCode evaluator
by default again and put it into the ROM-fs instead (so it still
can be loaded manually with "include evaluator.fs" if necessary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
No functional change in this patch, only cosmetics:
Use indentation width = 4 spaces everywhere, fix typo in
one of the comments ("Child pin" instead of "Chile pin"),
and rename "swizzledpin" in the return stack comments
to "parentpin" to match the non-return stack comments.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
PCI bridges can have up to 31 slots. If we limit the value
from pci-addr2dev with "modulo 4", all devices in slots 4
and higher won't work correctly since the interrupt-map
property then contains wrong values in this case. For
example, when QEMU is started with the following command
line, Linux is not able to use the balloon device:
qemu-system-ppc64 ... \
-device pci-bridge,bus=pci.0,id=bridge1,chassis_nr=1,addr=0x6 \
-device virtio-balloon-pci,id=balloon0,bus=bridge1,addr=5
To fix this issue, simply remove the bogus "4 mod" calculation.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1366953
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The legacy PCI interrupts in the device tree are enumerated from 1
to 4. However, the code in pci-gen-irq-map-one generates numbers
between 0 and 3 instead. This renders devices unusable by Linux
in case they should use IRQ 4. For example, when starting QEMU with
qemu-system-ppc64 ... \
-device pci-bridge,bus=pci.0,id=bridge1,chassis_nr=1,addr=0x6 \
-device virtio-balloon-pci,id=balloon0,bus=bridge1,addr=3
Linux can not use the balloon device since it is unable to determine
the right IRQ number in this case.
To fix this issue, we've simply got to make sure that the IRQ
numbers are in the range from 1 to 4 instead of 0 to 3.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1366953
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
When loading a file via the spapr-vlan network interface, SLOF
currently leaks quite a lot of memory, about 12kB each time.
This happens mainly because veth_init() uses for example
SLOF_alloc_mem_aligned(8192, 4096) and similar calls to get
the memory, but the space for the additional alignment is never
returned anymore later. An easy way to ease this situation is to
improve SLOF_alloc_mem_aligned() a little bit. We normally get memory
from SLOF_alloc_mem() which is aligned pretty well already, thanks to
the buddy allocator in SLOF. So we can try to first get a memory
block with the exact size first, and only do the alignment dance
if the first allocation was not aligned yet.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
It would put the pointer and len in the wrong order in the instance>args
buffer. As alloc-mem can handle zero length itself (and return NULL),
this also removes "if" to make the code simpler.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
[aik: removed "if" as Segher suggested; removed pipe from comment as Thomas suggested]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: Thomas Huth <thuth@redhat.com>
|
|
With my version of GCC (v4.8.5 - Advance-Toolchain 7.0) there are
currently two warnings when compiling sloffs.c:
sloffs.c: In function 'sloffs_dump':
sloffs.c:437:6: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
printf("%04x", be16_to_cpu(*(uint16_t *)(header->date + 2)));
^
sloffs.c:449:6: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
printf("%04x", be16_to_cpu(*(uint16_t *)(header->mdate + 2)));
^
These can be easily fixed by accessing the memory byte by byte instead of
casting the pointer to (uint16_t *). And while we're at it, let's also
simplify the code a little bit by consolidating the date print code
into a separate function which can be used by the two spots that print
a date.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The first two bytes of the date field are unused - and the date
is indeed encoded in the following six bytes, i.e. also using
the padding bytes. So the header definition is currently very
confusing. Let's simply remove the padding fields and make the
date field a little bit bigger instead - and add a proper comment
about the encoding of the date field.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Adrian Reber <areber@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Since this header file is not included from assembler files anymore,
we can remove the "#ifndef __ASSEMBLER__" here now and get rid of
the unnecessary indentation in this file.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Adrian Reber <areber@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The defines from calculatecrc.h are not used in any of the
assembler files anymore. So there's no need anymore to include
this file here.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Adrian Reber <areber@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Seems like they are completely unused, thus can be removed without
problems.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Adrian Reber <areber@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
When we will link the TFTP code to paflof instead of net-snk later,
the executable won't be reloaded each time we want to do an TFTP
transfer, so we got to make sure to initialize global variables
properly instead of relying on the ELF loader to do this job.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
With the addition of cpu hotplug in QEMU, cpu@0 can be removed as well.
SLOF should not depend on it. Find the first child in the "/cpus" node
and get the timer base frequency and set it as the chosen cpu as well
Reported-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
So far, the code was always assuming SuperSpeed for all devices,
which seemed to work OK with QEMU ... but let's better play safe
instead and use the speed from the port status instead.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
To support USB3 devices attached to a hub, we've got to set
the so-called "route string" in the slot context, which is
a concatenation of the hub port numbers that need to be
traversed between the root hub port and the destination device.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
USB3 devices need to be initialized with usb3_dev_init(), so that
they get assigned a proper slot ID. And for USB3 devices that are
attached to a (non-root) hub, we also need to keep track of the
hub topology, so a new field called "hub" is added to the struct
usb_dev which references the hub devices where the current USB
devices is attached to. The hub topology will be used later to
build the so-called "route string" for the USB3 devices.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
SLOF should be able to support more than just four devices on
the four root hub ports. So we've got to increase the amount of
possible slot IDs. Since QEMU supports up to 44 devices (four
devices at the root ports + 5 tiers of hubs with 8 ports each),
this seems to be a good new value instead. And to make sure that
we do not accidentially get a buffer overflow when accessing the
xhcd->xdevs[slot_id] array, this patch also add a sanity check at
the beginning of the xhci_alloc_dev() function here.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Arrays should not be declared in a header (to avoid that it gets
declared multiple times when the header is included more than once),
so the ps_array_usbX settings should reside in the .c file instead.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
sloffs can now copy a SLOF flash image from one file to
another. This is especially useful when the input is not
a regular file, but the slof_flash device. The -c, --copy
parameter copies only the actual SLOF flash image instead
of the whole device. This should result in a file, which
is identical to the file from the SLOF build process.
(cherry picked from commit 587eb035ac1fb17c3b8ab236bb06e793f08bb3ea)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Rewrote all code accessing the input file to not mmap it anymore,
because the crappy slof_flash driver does not support mmap but
only read/lseek.
Using read/lseek sloffs can now operate on a flash file image or
directly use the slof_flash device using the same code.
(cherry picked from commit 23fa80154b22b14147467160bd33ebc23b9275d1)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
[aik: squashed f2aa0523c]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
The -a, --append can now be used to append any file to
flash image filesystem. It is required that -o, --output
specifies an output file name.
(cherry picked from commit 339c6f4d17cae84526c7cba63928bf7769de5abe)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Added CRC check of the header and complete image
when dumping the header.
(cherry picked from commit 645bc2cbd550829d43b460663b998b9d4685a2d2)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
sloffs is a tool to get information about the flash images
created by the SLOF build process and it will also be able
to modify those flash images
(cherry picked from commit ec9f4acc3977a81289704fae0ad8bb62f3a93ab1)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
[aik: squashed e1e52adb5, e6369bb10b, 4d6f40c47d, 82532ec08; fixed warnings]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
This moves the CRC code to its own files without dependencies on
board specific parts of SLOF. That why it can also be used from
other parts of the code.
(cherry picked from commit e0c2a1d61a66309fd6f0cec7efadd7e988bd03ab)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
(cherry picked from commit 56b5525adf5473195f890b2e932f39f0d17fed49)
Cherry picked from https://lisas.de/~adrian/slof/slof.git/
Signed-off-by: Adrian Reber <adrian@lisas.de>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
Commit c8bd0cdce2bd ("Caps is not always shift") introduced
a global "ctrl" variable and two new cases to handle the shift
key in check_key_code() in usb-hid.c. However, one of these cases
uses the key code for the "," key and thus prevents that key
from working in the VGA console. The global ctrl variable that
is set by these two cases is never read again, so this is
apparently completely useless code, likely just an accidential
leftover from debugging the original issue. So let's simply
remove that bad code to get the comma key working again.
Fixes: c8bd0cdce2bd7c4498d61079553c89747656614d
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
On FAT32, the directory entry contains a new field providing the
top 16-bits of the cluster number. We didn't use it, thus reading
the wrong sectors when trying to access files or directories
beyond block 0x10000.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
[aik: removed unrelated empty lines]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
|
A bug crept in while doing the virtio 1.0 enablement in
commit 6e4d62c2 (virtio-net: enable virtio 1.0)
+ idx = virtio_modern16_to_cpu(&virtiodev, vq_rx.used->idx);
[...]
- vq_rx.avail->ring[vq_rx.avail->idx % vq_rx.size] = id - 1;
+ vq_rx.avail->ring[idx % vq_rx.size] = virtio_cpu_to_modern16(&virtiodev, id - 1);
sync();
- vq_rx.avail->idx += 1;
+ vq_rx.avail->idx = virtio_cpu_to_modern16(&virtiodev, idx + 1);
Should be using avail->idx in place of used->idx.
Signed-off-by: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
|
