A galois field multiplication in field 2 is like binary multiplication,
however instead of doing ordinary binary additions, xor's are performed.
So no carries are considered.
Implement all variants via helpers. s390_vec_sar() and s390_vec_shr()
will be reused later on.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Easy, we can reuse an existing gvec helper.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Implement it similar to VECTOR COUNT LEADING ZEROS.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
For 8/16, use the 32 bit variant and properly subtract the added
leading zero bits.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
To carry out the comparison, we can reuse the existing gvec comparison
function. In case the CC is to be computed, save the result vector
and compute the CC lazily. The result is a vector consisting of all 1's
for elements that matched and 0's for elements that didn't match.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Fairly easy to implement, we can make use of the existing CC helpers
cmps64 and cmpu64 - we siply have to sign extend the elements.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Time to introduce read_vec_element_i32 and write_vec_element_i32.
Take proper care of properly adding the carry. We can perform both
additions including the carry via tcg_gen_add2_i32().
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Similar to VECTOR AVERAGE but without sign extension.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Handle 32/64-bit elements via gvec expansion and the 8/16 bits via
ool helpers.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Easy, as we can reuse existing gvec helpers.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Similar to VECTOR ADD COMPUTE CARRY, however 128-bit handling only.
Courtesy of Richard H.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Only slightly ugly, perform two additions. At least it is only supported
for 128 bit elements.
Introduce gen_gvec128_4_i64() similar to gen_gvec128_3_i64().
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
128-bit handling courtesy of Richard H.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Introduce two types of fancy new helpers that will be reused a couple of
times
1. gen_gvec_fn_3: Call an existing tcg_gen_gvec_X function with 3
parameters, simplifying parameter passing
2. gen_gvec128_3_i64: Call a function that performs 128 bit calculations
using two 64 bit values per vector.
Luckily, for VECTOR ADD we already have everything we need.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Combine all variant in a single handler. As source and destination
have different element sizes, we can't use gvec expansion. Expand
manually. Also watch out for overlapping source and destination
registers. Use a safe evaluation order depending on the operation.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-33-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Very similar to VECTOR LOAD WITH LENGTH, just the opposite direction.
Properly probe write access before modifying memory.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-32-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Similar to VECTOR LOAD MULTIPLE, just the opposite direction. Probe
write access first.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-31-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
As we only store one element, there is nothing to consider regarding
exceptions.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-30-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Load both elements signed and store them into the two 64 bit elements.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-27-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Provide an implementation based on i64 and on real host vectors.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-26-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Similar to VECTOR GATHER ELEMENT, but the other direction.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-25-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Like VECTOR REPLICATE, but the element to be replicated comes from an
immediate.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-24-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Replicate via the special gvec helper.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-23-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Read the whole input before modifying the destination vector.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-22-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Take care of overlying inputs and outputs by using a temporary vector.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-21-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
This is a big one. Luckily we only have a limited set of such nasty
instructions.
We'll implement all variants with helpers, except when sources and
the destination don't overlap for VECTOR PACK. Provide different helpers
when the cc is to be modified. We'll return the cc then via env->cc_op.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-20-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
We cannot use gvec expansion as source and destination elements are
have different element numbers. So we'll expand using a fancy loop.
Also, we have to take care of overlapping source and destination
registers, therefore use a safe evaluation irder depending on the
operation.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-19-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
We can reuse the helper introduced along with VECTOR LOAD TO BLOCK
BOUNDARY. We just have to take care of converting the highest index into
a length.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-18-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Fairly easy, just load from to gprs into a single vector.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-17-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Very similar to VECTOR LOAD GR FROM VR ELEMENT, just the opposite
direction. Also provide a fast path in case we don't care about the
register content.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-16-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Very similar to LOAD COUNT TO BLOCK BOUNDARY, but instead of only
calculating, the actual vector is loaded. Use a temporary vector to
not modify the real vector on exceptions. Initialize that one to zero,
to not leak any data. Provide a fast path if we're loading a full
vector.
As we don't have gvec ool handlers for single vectors, just calculate
the vector address manually.
We can reuse the helper later on for VECTOR LOAD WITH LENGTH. In fact,
we are going to name it "vll" right from the beginning, because that's
a better match.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-15-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Try to load the last element first. Access to the first element will
be checked afterwards. This way, we can guarantee that the vector is
not modified before we checked for all possible exceptions. (16 vectors
cannot cross more than two pages)
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-14-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Fairly easy, zero out the vector before we load the desired element.
Load the element before touching the vector.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-13-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
To avoid an helper, we have to do the actual calculation of the element
address (offset in cpu_env + cpu_env) manually. Factor that out into
get_vec_element_ptr_i64(). The same logic will be reused for "VECTOR
LOAD VR ELEMENT FROM GR".
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-12-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Take care of properly sign-extending the immediate.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-11-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Fairly easy, load with desired size and store it into the right element.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-10-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
We can use tcg_gen_gvec_dup_i64() to carry out the duplication.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-9-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
When loading from memory, load both elements into temps first before
modifying the target vector
Loading with strange alingment from the end of the address space will
not properly wrap, we can ignore that for now.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-8-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Add gen_gvec_dupi() for handling duplication of immediates, so it can
be reused later.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-7-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Let's optimize it for the common cases (setting a vector to zero or all
ones) - courtesy of Richard.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-6-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Let's start with a more involved one, but it is the first in the list
of vector support instructions (introduced with the vector facility).
Good thing is, we need a lot of basic infrastructure for this. Reading
and writing vector elements as well as checking element validity.
All vector instruction related translation functions will reside in
translate_vx.inc.c, to be included in translate.c - similar to how
other architectures handle it.
While at it, directly add some documentation (which contains parts about
things added in follow-up patches, but splitting this up does not make
too much sense). Also add ES_* defines heavily used later.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190307121539.12842-5-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
With the floating-point extension facility, LOAD ROUNDED has
a rounding mode specification and the inexact-exception control (XxC).
Handle them just like e.g. LOAD FP INTEGER.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190218122710.23639-14-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Let's split handling of BFP/DFP rounding mode configuration. Also,
let's not reuse the sfpc handler, use a separate handler so we can
properly check for specification exceptions for SRNMB.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190218122710.23639-10-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Use a new CC helper to calculate the CC lazily if needed. While the
PoP mentions that "A 32-bit unsigned binary integer" is placed into the
first operand, there is no word telling that the other 32 bits (high
part) are left untouched. Maybe the other 32-bit are unpredictable.
So store 64 bit for now.
Bit magic courtesy of Richard.
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190225200318.16102-8-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Nice trick to load a 32 bit value into vector element 0 (32 bit element
size) from memory, zeroing out element1. The short HFP to long HFP
conversion really only is a shift.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190225200318.16102-7-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
This is a non-privileged instruction that was only implemented
for system mode. However, the stck instruction is used by glibc,
so this was causing SIGILL for programs run under debian stretch.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20190212053044.29015-3-richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
As floating point registers overlay some vector registers and we want
to make use of the general tcg_gvec infrastructure that assumes vectors
are not stored in globals but in memory, don't model floating point
registers as globals anymore. This is then similar to how arm handles
it.
Reading/writing a floating point register means reading/writing memory now.
Break up ugly in2_x2() handling that modifies both, in1 and in2 into
in2_x2l and in2_x2h. This makes things more readable. Also, in1_x1() is
ugly as it touches out/out2, get rid of that and use prep_x1() instead.
As we are no longer able to use the original global variables for
out/out2, we have to use new temporary variables and write from them to
the target registers using wout_ helpers.
E.g. an instruction that reads and writes x1 will use
- prep_x1 to get the values into out/out2
- wout_x1 to write the values from out/out2
This special handling is needed for x1 as it is often used along with
other inputs, so in1/in2 is already used.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20190204154406.16122-1-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Let's check this also at a central place.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20180927130303.12236-8-david@redhat.com>
Acked-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
These flags allow us to later on detect if a DATA program interrupt
is to be injected, and which DXC (1,2,3) is to be used.
Interestingly, some support FP instructions are considered as HFP
instructions (I assume simply because they were available very early).
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20180927130303.12236-6-david@redhat.com>
Acked-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
David Hildenbrand