Brian Gaeke [Thu, 17 Jul 2003 18:53:20 +0000 (18:53 +0000)]
Added bits about MachineFunctionPass
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7192
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:48:38 +0000 (21:48 +0000)]
Tests for globals with different kinds of behavior in DS Analysis.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7191
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:45:15 +0000 (21:45 +0000)]
(1) Added DSGraph::cloneReachableSubgraph and DSGraph::cloneReachableNodes
to clone the subgraph reachable from a set of root nodes, into the
current graph, merging the global nodes into those in the current graph.
(2) Added DSGraph::updateFromGlobalGraph() to rematerialize nodes from the
globals graph into the current graph in both BU and TD passes.
(3) Added hash_set<const GlobalValue*> InlinedGlobals: a set of globals to
track which globals have been inlined into the current graph from
callers or callees. In the TD pass, such globals are up-to-date and
do not need to be rematerialized from the GlobalsGraph.
(4) Added StripIncompleteBit/KeepIncompleteBit to remove incomplete bit
when cloning nodes into the globals graph.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7190
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:42:03 +0000 (21:42 +0000)]
Rematerialize nodes from the globals graph into the current graph
after all callees are inlined into the current graph.
NOTE: There's also a major bug fix for the BU pass in DataStructure.cpp,
which ensures that resolvable indirect calls are not moved out to the
globals graph, so that they are eventually inlined (if possible).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7189
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:40:28 +0000 (21:40 +0000)]
(1) Rematerialize nodes from the globals graph into the current graph
after all callers are inlined into the current graph.
(2) Optimize the way a graph is inlined into its callees in the TD phase:
(a) Use DSGraph::cloneReachableSubgraph to clone only a subgraph at
each call site, for faster inlining.
(b) Clone separately for the same callee at different call sites,
since only the reachable subgraph is being cloned, not the entire
caller graph.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7188
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:36:31 +0000 (21:36 +0000)]
Implement 2 important changes: (1) rematerialization from the globals graph,
and (2) faster inlining by cloning only reachable nodes. In particular:
(1) Added DSGraph::cloneReachableSubgraph and DSGraph::cloneReachableNodes
to clone the subgraph reachable from a set of root nodes, into the
current graph, merging the global nodes into thos in the current graph.
The TD pass now uses this for faster inlining, and so does the
next function.
(2) Added DSGraph::updateFromGlobalGraph() to rematerialize nodes from the
globals graph into the current graph in both BU and TD passes.
(3) `I' flags are removed from all nodes in the globals graph, because they
are difficult to maintain correctly and are not needed anyway.
(4) Aux. function calls are only removed to the globals graph if they
will never be resovled. (This is what fixed gap.) The immediate
reason is that if we took these out of a function (and moved them to
the globals graph) we would need to rematerialize these nodes into the
function graph for every function in the BU pass. The longer term
problem is that we would need to find a way to remove them from the
globals graph iff they have been resolved on all paths through the
call graph.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7187
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 16 Jul 2003 21:25:17 +0000 (21:25 +0000)]
Factor out the test for unresolvable external functions into
isUnresolvableFunc() (I thought I needed this externally.
I don't, but it's still nicer this way.)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7186
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 16 Jul 2003 20:30:40 +0000 (20:30 +0000)]
Fixed the number translation scheme for the integer condition code registers: it
now works in instructions which require a 2-bit or 3-bit INTcc code.
Incidentally, that means that the representation of INTcc registers is now the
same in both integer and FP instructions. Thus, code became much simpler and
cleaner.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7185
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 16 Jul 2003 20:27:44 +0000 (20:27 +0000)]
The name should really be `simm11' to follow the naming convention, but this has
no change in functionality.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7184
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Wed, 16 Jul 2003 20:26:06 +0000 (20:26 +0000)]
Marked some of the phony targets are PHONY. This will hopefully speed
up builds a little bit since it will prevent GNU make from matching these
phony targets against implicit rules.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7183
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 21:27:14 +0000 (21:27 +0000)]
No need for a second immediate field if the class already inherits one.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7182
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 21:26:49 +0000 (21:26 +0000)]
Encode predict = 1 by default, because the Sparc assembler does this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7181
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 21:26:09 +0000 (21:26 +0000)]
Fixed a bug: outputting name of variable instead of its value.
Also, placed DEBUG() guards around debug information so that the generated file
is much smaller and hence should be faster to preprocess/compile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7180
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 21:00:32 +0000 (21:00 +0000)]
This optimization greatly enhances efficiency of creating new instructions by
masking and shifting operands directly into their place in the instruction,
instead of the old-fashioned way of ORing in each bit separately.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7179
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 19:09:43 +0000 (19:09 +0000)]
Correctly handle calls to functions which are further away than 2**32 bits will
allow, i.e. make a sequence of instructions to enable an indirect call using
jump-and-link and 2 temporary registers (which we save and ultimately restore).
Warning: if the delay slot of a function call is used to do meaningful work and
not just a NOP, this behavior is incorrect. However, the Sparc backend does not
yet utilize the delay slots effectively, so it is not necessary to make an
overly complicated algorithm for something that's not used.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7178
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 15:58:26 +0000 (15:58 +0000)]
Clean up my last checkin: code is easier to read and explains the differences in
usage of the special file handle RTLD_SELF on Sparc/Solaris vs. 0 on Linux/x86.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7177
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 15 Jul 2003 15:55:32 +0000 (15:55 +0000)]
On Sparc/Solaris, the special handle RTLD_SELF is used as a handle referring to
the program's executing image, not 0 as it is on Linux/x86 and possibly other
systems.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7176
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 14 Jul 2003 23:26:03 +0000 (23:26 +0000)]
* Added support for the %ccr register
* FP double registers are now coded correctly
* Removed function which converted registers based on register types, it was
broken (because regTypes are broken)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7175
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 14 Jul 2003 17:26:34 +0000 (17:26 +0000)]
The word `separate' only has one `e'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7174
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 14 Jul 2003 17:20:40 +0000 (17:20 +0000)]
The word `separate' only has one `e'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7173
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 14 Jul 2003 17:05:20 +0000 (17:05 +0000)]
Fixed: </pre</p> => </pre></p>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7172
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Mon, 14 Jul 2003 16:52:07 +0000 (16:52 +0000)]
Merged in changes from PRE101 in the prerelease branch.
Specifically, this updates libtool to version 1.5 and adds the following:
- Added the -only-static option that we added in our previous libtool.
- Modified the autoconf macros so that libtool uses the -G option when
linking on Solaris. This allows libraries with global variables with
constructors to automatically run those constructors when the
library is dlopened().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7171
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Fri, 11 Jul 2003 22:02:28 +0000 (22:02 +0000)]
Use uint32_t for table index and size: table will never be > 4GB.
Also, make Pointer type depend on architecture.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7170
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Fri, 11 Jul 2003 21:57:43 +0000 (21:57 +0000)]
Trace loads and stores as they happen (stores were being
remembered in valuesStoredInFunction, but never traced at function return,
and that's too late to be finding the error anyway).
Stores trace both the value and the address being stored to,
but after some experience I think only values should be traced.
The pointer hash table just fills up far too quickly if every
store address were traced.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7169
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Fri, 11 Jul 2003 21:57:01 +0000 (21:57 +0000)]
If we have a constant pointer reference to a function, we were printing
out the entire llvm disassembly for the function at global constant-output
time, which caused the assembler to barf in 164.gzip. This fixes that
particular problem (though 164.gzip has other problems with X86 llc.)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7168
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Fri, 11 Jul 2003 18:18:35 +0000 (18:18 +0000)]
Fix some serious floating-point bugs (fixes test cases such as Oscar,
Fhourstones, McCat-vor, and many others...)
Printer.cpp: Print implicit uses for AddRegFrm instructions. Break gas
bug workarounds up into separate stanzas of code for each bug. Add new
workarounds for fild and fistp.
X86InstrInfo.def: Add O_ST0 implicit uses for more FP instrs where they
obviously apply. Also add PrintImplUses flags for FP instrs where they
are necessary for gas to understand the output.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7165
91177308-0d34-0410-b5e6-
96231b3b80d8
Anand Shukla [Thu, 10 Jul 2003 21:55:57 +0000 (21:55 +0000)]
Added functionality to instrmentation pass
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7161
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 10 Jul 2003 21:38:28 +0000 (21:38 +0000)]
Fixed grammatical error.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7160
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 20:07:54 +0000 (20:07 +0000)]
Several important bug fixes:
(1) Cannot use ANDN(ot), ORN, and XORN for boolean ops, only bitwise ops.
(2) Conditional move instructions must distinguish signed and unsigned
condition codes, e.g., MOVLE vs. MOVLEU.
(3) Conditional-move-on-register was using the cond-move-on-cc opcodes,
which produces a valid-looking instruction with bogus registers!
(4) Here's a really cute one: dividing-by-2^k for negative numbers needs to
add 2^k-1 before shifting, not add 1 after shifting. Sadly, these
are the same when k=0 so our poor test case worked fine.
(5) Casting between signed and unsigned values was not correct:
completely reimplemented.
(6) Zero-extension on unsigned values was bogus: I was only doing the
SRL and not the SLLX before it. Don't know WHAT I was thinking!
(7) And the most important class of changes: Sign-extensions on signed values.
Signed values are not sign-extended after ordinary operations,
so they must be sign-extended before the following cases:
-- passing to an external or unknown function
-- returning from a function
-- using as operand 2 of DIV or REM
-- using as either operand of condition-code setting operation
(currently only SUBCC), with smaller than 32-bit operands
Also, a couple of improvements:
(1) Fold cast-to-bool into Not(bool). Need to do this for And, Or, XOR also.
(2) Convert SetCC-Const into a conditional-move-on-register (case 41)
if the constant is 0. This was only being done for branch-on-SetCC-Const
when the branch is folded with the SetCC-Const.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7159
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:48:19 +0000 (19:48 +0000)]
Bug fix in creating constants: need 1U << 31, not 1 << 31.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7158
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:47:42 +0000 (19:47 +0000)]
Fold cast-to-bool into not. Later, this should also be folded into other
boolean operations: AND, OR, XOR.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7157
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:46:15 +0000 (19:46 +0000)]
Values stored in CallArgsDescriptor cannot be const.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7156
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:45:28 +0000 (19:45 +0000)]
isMarkedForSpill() should be const.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7155
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:45:07 +0000 (19:45 +0000)]
Change interface to MachineInstr::substituteValue to specify more precisely
which args can be substituted: defsOnly, defsAndUses or usesOnly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7154
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:43:33 +0000 (19:43 +0000)]
Print out all neighbors in interference graph, not just higher-numbered ones.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7153
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:42:55 +0000 (19:42 +0000)]
Several fixes to handling of int CC register:
(1) An int CC live range must be spilled if there are any interferences,
even if no other "neighbour" in the interf. graph has been allocated
that reg. yet. This is actually true of any class with only one reg!
(2) SparcIntCCRegClass::colorIGNode sets the color even if the LR must
be spilled so that the machine-independent spill code doesn't have to
make the machine-dependent decision of which CC name to use based on
operand type: %xcc or %icc. (These are two halves of the same
register.)
(3) LR->isMarkedForSpill() is no longer the same as LR->hasColor().
These should never have been the same, and this is necessary now for #2.
(4) All RDCCR and WRCCR instructions are directly generated with the
phony number for %ccr so that EmitAssembly/EmitBinary doesn't have to
deal with this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7152
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:42:11 +0000 (19:42 +0000)]
Several fixes to handling of int CC register:
(1) An int CC live range must be spilled if there are any interferences,
even if no other "neighbour" in the interf. graph has been allocated
that reg. yet. This is actually true of any class with only one reg!
(2) SparcIntCCRegClass::colorIGNode sets the color even if the LR must
be spilled so that the machine-independent spill code doesn't have to
make the machine-dependent decision of which CC name to use based on
operand type: %xcc or %icc. (These are two halves of the same register.)
(3) LR->isMarkedForSpill() is no longer the same as LR->hasColor().
These should never have been the same, and this is necessary now for #2.
(4) All RDCCR and WRCCR instructions are directly generated with the
phony number for %ccr so that EmitAssembly/EmitBinary doesn't have to
deal with this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7151
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:32:38 +0000 (19:32 +0000)]
*** empty log message ***
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7150
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:31:26 +0000 (19:31 +0000)]
Just documentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7149
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Thu, 10 Jul 2003 19:25:29 +0000 (19:25 +0000)]
Add phony target "bytecode" to ensure that it works in recursive makes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7147
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 10 Jul 2003 17:05:26 +0000 (17:05 +0000)]
Lowercase versions of `occurrence' need to be spelled correctly, too.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7142
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 10 Jul 2003 16:52:41 +0000 (16:52 +0000)]
Move the space separator to where it really belongs: in the print statement, not
in a variable assignment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7141
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 10 Jul 2003 16:49:51 +0000 (16:49 +0000)]
`Occurrence' has no `a' and the `r' is doubled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7140
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 10 Jul 2003 16:35:17 +0000 (16:35 +0000)]
Converted tabs to spaces.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7139
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 9 Jul 2003 17:33:50 +0000 (17:33 +0000)]
Some beautification changes (tabs to spaces, removed extra blank lines);
no functionality changed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7138
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Tue, 8 Jul 2003 23:40:48 +0000 (23:40 +0000)]
Simplified DOTRACING flag.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7135
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Tue, 8 Jul 2003 20:35:59 +0000 (20:35 +0000)]
Corrected the Table of Contents.
Corrected capitalization of subheadings.
Created a new subsection for compiling the C front end and moved all references
to it there.
Updated the disk space requirements to reflect the need for the binary C front
end and the optional C front end source.
Added information on unpacking the distribution to the summary section.
Moved autoconf environment variables to the autoconf section.
Changed make to gmake.
Removed some of the precise directions for unpacking the archives.
Fixed some formatting inconsistencies (headings that were not centered).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7132
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Tue, 8 Jul 2003 18:42:44 +0000 (18:42 +0000)]
Pointer hash table reallocation code seems never to have been tested!
Unfortunately, reallocation also means that the pointer numbering will
change, so increase table size to try to avoid it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7130
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Tue, 8 Jul 2003 18:40:58 +0000 (18:40 +0000)]
Native libraries (libinstr*.a) should not have been taken out when
taking out the rule for compiling the test driver.
We need the native libraries for libinstr because they are directly
linked into the native CBE or LLC code. That is the only practical
way to debug them!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7129
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Tue, 8 Jul 2003 18:39:51 +0000 (18:39 +0000)]
Use libinstr.$(ARCH).a instead of libinstr64.a
Also, $(RM) needs -f flag.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7128
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Tue, 8 Jul 2003 15:39:02 +0000 (15:39 +0000)]
*** empty log message ***
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7126
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Tue, 8 Jul 2003 14:52:09 +0000 (14:52 +0000)]
Removed size_t from the parameter list of the malloc() prototype in generated C
code. This prevents a compiler warning on Sparc that causes the tests to fail.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7125
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 22:30:44 +0000 (22:30 +0000)]
Stop using the `Offset' variable, as we are cycling through the bits of a field
initializer and the loop index variable already carries the offset information
that we need.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7123
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 22:28:42 +0000 (22:28 +0000)]
Fixed indentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7122
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 22:27:05 +0000 (22:27 +0000)]
Removed an extra slash that appears in the path name when these variables are
combined with a '/' separating them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7121
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 22:18:42 +0000 (22:18 +0000)]
Elaborated assembly syntax of instructions in the comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7120
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 22:18:06 +0000 (22:18 +0000)]
Removed unnecessary assignment (it was taken care by a superclass) and clarified
some comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7119
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 21:27:40 +0000 (21:27 +0000)]
* Made $Template refer to a path relative to the checked-out tree
* Made error messages print out what directory failed chdir()
* Put quotes around search arguments to grep
* Use `egrep' instead of `grep -e' because they are equivalent but `grep' does
not have the `-e' option on Sparc/Solaris
* Added `--enable-jit' to the ./configure command because both X86 and Sparc
have JITs and we want them to be tested
* Fixed the regular expressions parsing the changes in CVS which were causing
the script to die
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7118
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Mon, 7 Jul 2003 19:29:20 +0000 (19:29 +0000)]
Fixed the directions for building the C front end.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7117
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Mon, 7 Jul 2003 19:27:35 +0000 (19:27 +0000)]
Added information about how to unpack the distribution for those who do not
have access to CVS.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7116
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Mon, 7 Jul 2003 18:34:20 +0000 (18:34 +0000)]
Insert workaround for GAS bug in assembling FLD/FSTP XWORD PTR [...]
instructions, by outputting them as bytes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7115
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 7 Jul 2003 16:52:39 +0000 (16:52 +0000)]
Moved RegClassIDs enum to be next to the RegTypes enum.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7114
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Sun, 6 Jul 2003 22:50:31 +0000 (22:50 +0000)]
Correction to last fix: Pointer types do not return true in Type::IsIntegral().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7113
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Sun, 6 Jul 2003 20:33:21 +0000 (20:33 +0000)]
Choose register instead of immediate for ConstantExpr in ChooseRegOrImmed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7112
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Sun, 6 Jul 2003 20:13:59 +0000 (20:13 +0000)]
Major bug fix though it happened rarely (only on a compare after an
integer overflow):
We need to use %icc and not %xcc for comparisons on 32-bit or smaller
integer values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7111
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Sun, 6 Jul 2003 20:12:18 +0000 (20:12 +0000)]
Regression test for bug in reg. allocation that was using %xcc instead
of %icc even for 32-bit and smaller comparisons.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7110
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Sun, 6 Jul 2003 19:53:59 +0000 (19:53 +0000)]
Make the RegClassID values public -- there is no other way to get them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7109
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Thu, 3 Jul 2003 21:44:32 +0000 (21:44 +0000)]
Add -d option to trust the disassembler.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7105
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Thu, 3 Jul 2003 21:15:02 +0000 (21:15 +0000)]
here little scriptie, nice scriptie...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7104
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 3 Jul 2003 18:36:47 +0000 (18:36 +0000)]
Apparently, the "regType" and "regClass" used in the Sparc backend are not both
correct: empirically, "regType" is wrong for a number of registers. Thus, one
can only rely on the "regClass" to figure out what kind of register one is
dealing with.
This change switches to using only "regClass" and adds a few extra DEBUG() print
statements and a few clean-ups in comments and code, mostly minor.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7103
91177308-0d34-0410-b5e6-
96231b3b80d8
Sumant Kowshik [Thu, 3 Jul 2003 17:55:47 +0000 (17:55 +0000)]
Added support for poolallocarray and poolmakeunfreeable. The latter is used by the SAFECode project
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7102
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Thu, 3 Jul 2003 16:49:40 +0000 (16:49 +0000)]
Fixed an error in the pathname to LLVMGCCDIR for the pre-built binaries (forgot
that <> denotes a tag).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7100
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Thu, 3 Jul 2003 16:43:01 +0000 (16:43 +0000)]
Added information on software requirements for unpacking the archives that we
will eventually distribute.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7099
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Thu, 3 Jul 2003 16:29:36 +0000 (16:29 +0000)]
* Reworded a bit about JITs (I don't like acronyms followed by 's)
* Added JIT capability to the LLI description section
* LLC is quasi-x86-capable
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7098
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Thu, 3 Jul 2003 16:03:39 +0000 (16:03 +0000)]
Added another remark on how building the C front end is optional.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7097
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Thu, 3 Jul 2003 16:01:38 +0000 (16:01 +0000)]
Performed a major update for the pre-release.
Attempted to better structure and order the document.
Added more information about autoconf, the build system, and how to build LLVM.
Added directions on how to build the C front end.
Added a section on common problems and their solutions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7096
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Thu, 3 Jul 2003 15:37:52 +0000 (15:37 +0000)]
Adding a web page on how to start a new LLVM Project.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7095
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 3 Jul 2003 06:42:38 +0000 (06:42 +0000)]
Reuse the values if they are constants: this is important so that we index into the right structure field
This fixes bug: BasicAA/2003-07-03-BasicAACrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7093
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 3 Jul 2003 06:29:15 +0000 (06:29 +0000)]
New testcase
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7092
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 3 Jul 2003 02:03:53 +0000 (02:03 +0000)]
Remove globals more aggressively from graphs.
Fix a bug where we removed nodes that were marked U.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7090
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:57:21 +0000 (23:57 +0000)]
INCLUDE_PARENT_GRAPH is required
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7089
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:57:05 +0000 (23:57 +0000)]
INCLUDE_PARENT_GRAPH is required!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7088
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:56:51 +0000 (23:56 +0000)]
Disable incorrect mustalias code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7087
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:44:15 +0000 (23:44 +0000)]
Remove space at end of line
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7084
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:43:06 +0000 (23:43 +0000)]
Remove dead method
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7083
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 23:42:48 +0000 (23:42 +0000)]
Fix how we are handling unreachable functions. This DRAMATICALLY improves efficiency
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7082
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Wed, 2 Jul 2003 23:03:18 +0000 (23:03 +0000)]
Updated prior to the pre-release.
Removed items which are done (or near completion).
Added new items so that we don't forget them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7081
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 21:20:04 +0000 (21:20 +0000)]
The word `open' was really intended to be `option'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7080
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Wed, 2 Jul 2003 21:16:06 +0000 (21:16 +0000)]
Removed the --enable-profiling option.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7079
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Wed, 2 Jul 2003 20:49:38 +0000 (20:49 +0000)]
Removed the --enable-profiling option (again).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7078
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 20:24:42 +0000 (20:24 +0000)]
Keep track of how many inlinings are performed
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7076
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 19:49:11 +0000 (19:49 +0000)]
Try using trivially dead deletion
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7075
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 19:37:48 +0000 (19:37 +0000)]
* Force all "don't care" bits to 0 so that there are absolutely no unset bits in
the TableGen descriptions; all unset bits are thus errors.
* As a result, found and fixed instructions where some operands were not
actually assigned into the right portion of the instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7074
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 18:27:47 +0000 (18:27 +0000)]
The classes F4_3 and F4_4 have an `rd' operand that needs to be set.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7073
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 18:15:43 +0000 (18:15 +0000)]
Properly fix instruction syntax in comments, using `imm' for instructions that
use an immediate value instead of a register.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7072
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 18:02:58 +0000 (18:02 +0000)]
Fixed instruction syntax in the comments (specifies how instr is used).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7071
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 2 Jul 2003 17:53:19 +0000 (17:53 +0000)]
* If compiling on X86 or Sparc, automagically enable the JIT for that arch
* Setting ENABLE_X86_JIT or ENABLE_SPARC_JIT on the `make' command-line will
force the inclusion of that JIT on a different architecture
* If neither JIT is enabled (e.g., compiling on a different architecture), the
-march option will not be available to LLI.
* As a side effect of the $ARCH variable, the Sparc LLI can now link just a bit
faster by not including the x86 library.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7070
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Wed, 2 Jul 2003 16:53:44 +0000 (16:53 +0000)]
Removed the CPPFLAGS and CFLAGS variables since the -DHAVE_CONFIG_H messes
up the test suite. Since all the LLVM software assumes that config.h exists,
we can just do away with it for now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7069
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram S. Adve [Wed, 2 Jul 2003 06:59:22 +0000 (06:59 +0000)]
Force fixed-size but large alloca objects to the dynamically allocated
area to avoid using up precious stack space within the 4095 offset limit
from %fp. Such objects that would themselves live at a large offset
were being put there already so this is a simple change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7066
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 2 Jul 2003 06:06:34 +0000 (06:06 +0000)]
Remove dead Nodes list
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7065
91177308-0d34-0410-b5e6-
96231b3b80d8