This patch contains several small cleanups:
- In qcom_smem_enumerate_partitions(), change the "local_host"
argument to have 16 bit unsigned type
- Also in qcom_smem_enumerate_partitions(), change the type of
the "host0" and "host1" local variables to be u16
- Fix error messages reporting host ids to use the right format
specifier
- Shorten the error messages as well, to fit on one line
- Add a compile-time check to ensure the local host value passed
to qcom_smem_enumerate_partitions() is in range
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Add verification in qcom_smem_partition_header() that the host ids
found in a partition's header structure match those in its partition
table entry.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Change the logic in the loop that finds that global host entry in
the partition table not require the host0 and host1 local variables.
The next patch will remove them.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Add verification in qcom_smem_partition_header() that the
offset_free_uncached field in a partition's header structure does
not exceed the partition's size.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Add verification in qcom_smem_partition_header() that the size in a
partition's header structure matches the size in its partition table
entry.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Create a new function qcom_smem_partition_header() to encapsulate
validating locating a partition header and validating information
found within it. This will be built up over a few commits to make
it more obvious how the common function is replacing duplicated code
elsewhere. Initially it just verifies the header has the right
magic number.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
In qcom_smem_enumerate_partitions(), we find all partitions that
have a given local host id in either its host0 or its host1 field
in the partition table entry. We then verify that the header
structure at the start of each partition also contains the same two
host ids as is found in the table of contents.
There is no requirement that the order of the two host ids be the
same in the table of contents and in the partition header.
This patch changes that, requiring host0 to in the partition table
entry to equal host0 in the partition header structure (and similar
for the host1 values).
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
The global partition is indicated by having both host values in its
table of contents entry equal SMEM_GLOBAL_HOST=0xfffe.
In qcom_smem_set_global_partition(), we check whether the header
structure at the beginning of the partition contains that host
value, but the check only verifies *one* of them. Change the check
so the partition header must have SMEM_GLOBAL_HOST for *both* its
host fields.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Combine the code that checks whether a partition table entry is
associated with the local host with the assignment of the remote
host id value.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
In qcom_smem_enumerate_partitions(), any partition table entry
having a zero offset or size field is ignored. Move those checks
earlier in the loop, because there's no sense in examining the
host fields for those entries.
Add the same checks in qcom_smem_set_global_partition(), so the
scan for the global partition skips over these invalid entries.
This allows a later check for zero size or offset once the global
entry is found to be eliminated.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Hold off initializing anything for the array entry representing a
memory region in qcom_smem_map_memory() until we know we've
successfully mapped it.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Rename the variable "area" to be "region" in qcom_smem_get_global(),
so its name better matches its type.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
The moved check for the global partition ended up in the wrong place and I
failed to spot this in my review. This moves it to the correct place.
Fixes: 11d2e7edac ("soc: qcom: smem: check sooner in qcom_smem_set_global_partition()")
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Create function qcom_smem_virt_to_phys(), which returns the physical
address corresponding to a given SMEM item's virtual address. This
feature is required for a driver that will soon be out for review.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
There's no sense in scanning the partition table again if we know
the global partition has already been discovered. Check for a
non-null global_partition pointer in qcom_smem_set_global_partition()
immediately.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
If there is at least one entry in the partition table, but no global
entry, the qcom_smem_set_global_partition() should return an error
just like it does if there are no partition table entries.
It turns out the function still returns an error in this case, but
it waits to do so until it has mistakenly treated the last entry in
the table as if it were the global entry found.
Fix the function to return immediately if no global entry is found
in the table.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
It's OK if the space for a newly-allocated uncached entry actually
touches the free cached space boundary. It's only a problem if it
would cross it.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Two places report an error when a partition header is found to
not contain the right canary value. The error messages do not
properly byte swap the host ids. Fix this, and adjust the format
specificier to match the 16-bit unsigned data type.
Move the error handling for a bad canary value to the end of
qcom_smem_alloc_private(). This avoids some long lines, and
reduces the distraction of handling this unexpected problem.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
What phdr_to_last_uncached_entry() returns is the address of the
start of the free space following all allocated uncached entries.
It really doesn't refer to an actual (initialized) private entry
structure. Similarly phdr_to_last_cached_entry() returns the
address of the end of free space, preceding the last allocated cache
entry. Change both functions' return type to be pointer to void
to reflect this.
Meanwhile, phdr_to_first_cached_entry() really *does* point to a
private entry structure, so change its return type to reflect
this fact.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Cached items are found at the high end of an smem partition. A
cached item's shared memory precedes the private entry structure
that describes it.
The address of the structure describing the first cached item should
be returned by phdr_to_first_cached_entry(). However the function
calculates the start address using the wrong structure size.
Fix this by computing the first item's entry structure address by
subtracting the size of a private entry structure rather than a
partition header structure.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Increase the maximum number of hosts in a system to 10.
Signed-off-by: Chris Lew <clew@codeaurora.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
In V12 SMEM, SBL writes SMEM parameter information after the TOC. Use
the SBL provided item count as the max item number.
Signed-off-by: Chris Lew <clew@codeaurora.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
SMEM V12 creates a global partition to allocate global smem items from
instead of a global heap. The global partition has the same structure as
a private partition.
Signed-off-by: Chris Lew <clew@codeaurora.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
The SMEM header structure includes the version information. Read the
version directly from the header instead of getting an item from the
global heap.
Signed-off-by: Chris Lew <clew@codeaurora.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Endianness can vary in the system, add le32_to_cpu when comparing
partition sizes from smem.
Signed-off-by: Chris Lew <clew@codeaurora.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
On msm8996 cached SMEM items are used for storing the GLINK FIFOs, so
for items not found in the uncached list we need to also search the
cased list for these items.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
In preparation for adding accessors for "cached" entries rename the
"uncached" accessors. Also rename "first" cached entry to "last", as
the cached list grows backwards.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
If we fail to get the hwspinlock due to probe defer, we shouldn't
print an error message. Just be silent in this case.
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
Mapping the SMEM region as write combine makes the contiguous writes
in SMD perform better and also allows us to do unaligned read and writes
on ARM64.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Reviewed-by: Andy Gross <andy.gross@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
SMEM is a software construct built on top of a DDR reserved region
and sometimes a device memory region called RPM message ram. Having
the RPM message ram in the smem DT node's reg property leads to the
smem node being located in different places depending on if the
message ram is being used or not. Let's add a qcom specific
property, qcom,rpm-msg-ram, and point to the device memory from
the SMEM node via a phandle. As SMEM is a software construct, it
really needs to reside at the root of the DT regardless of whether
it's using the message ram or not.
Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Andy Gross <agross@codeaurora.org>
The contents of smem are always in little endian, but the smem
driver is not capable of being used on big endian CPUs. Annotate
the little endian data members and update the code to do the
proper byte swapping.
Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Andy Gross <agross@codeaurora.org>
Passing a void ** almost always requires a cast at the call site.
Instead of littering the code with casts every time this function
is called, have qcom_smem_get() return a void pointer to the
location of the smem item. This frees the caller from having to
cast the pointer.
Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Andy Gross <agross@codeaurora.org>
Don't set a pointer to NULL and then dereference it in the next
line.
Cc: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Andy Gross <agross@codeaurora.org>
This patch corrects private partition item access. Instead of falling back to
global for instances where we have an actual host and remote partition existing,
return the results of the private lookup.
Signed-off-by: Andy Gross <agross@codeaurora.org>
The Shared Memory Manager driver implements an interface for allocating
and accessing items in the memory area shared among all of the
processors in a Qualcomm platform.
Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Acked-by: Andy Gross <agross@codeaurora.org>
Signed-off-by: Andy Gross <agross@codeaurora.org>
Alex Elder