elfutils-devel November 2010

elfutils-devel@lists.fedorahosted.org

8 participants
10 discussions

dwarf_output overview
by Roland McGrath 05 Jan '11

05 Jan '11

https://fedorahosted.org/elfutils/wiki/DwarfOutput is still accurate in its basic description about the collector (class dwarf_output_collector). We don't really need to get into right now why it's a separate object from dwarf_output. For uses in the time being, we'll use one collector with one dwarf_output for one input file. The simple attribute values, and some bits of interface glue, are shared via template use from dwarf_data, with dwarf_edit. The only difference is that in dwarf_output we never have mutable versions of thse objects, only const pointers/references to immutable copies sitting in the collector. Whereas dwarf_edit has mutable containers, there is no modification you can do to a dwarf_output object once it's created. Instead, like dwarf_edit you can copy-construct a dwarf_output object from any dwarf-compatible class (i.e. dwarf, dwarf_edit, or dwarf_output)--this is the only way to get a nonempty dwarf_output object. The dwarf_output constructor takes two arguments, unlike the simple dwarf_edit constructor. The second argument is the collector object. This actually uses a second helper object, whose only purpose is to live as long as copy construction is going on. That's dwarf_output::copier, which is actually dwarf_output::copier<input> parameterized by the particular top-level input type (dwarf or dwarf_edit). (This will one day be exposed outside dwarf_output for some of the same cases where a separate collector is really useful, but we won't get into that now.) The copy construction is doing the whole file, or at least a whole CU, from the root of the tree down to the leaves and all attribute values. The copier holds some bookkeeping and some caches during that process. The essential method is that the tree from the CU on down has its debug_info_entry objects constructed recursively, with debug_info_entry::attr_value objects being constructed for each node. Each attr_value on up to each debug_info_entry is "constructed" by finding its exact match stored in the collector, inserting it only if it's not already there, and then returning the pointer (i.e. actually all C++ references) to the (now) existing data structure. The copier's part in this is twofold. First, there is some simple caching related to copying the simple attribute values. (All the value types other than references are "simple".) Second, there is all the hairy bookkeeping for the complexity brought on by references. The copier's caching is a simple optimization, so I'll describe it first. This is for strings, identifier strings (which we distinguish from other strings, though they are just strings in DWARF), and source files. In one input file, even across many CUs (because of linker SHF_MERGE|SHF_STRINGS merging), every DW_AT_name "foobar" will point to point to the same "foobar" in the .debug_str table. Likewise, a dwarf::source_file object that represents a source file from the .debug_line table is what we get for every DW_AT_decl_file attribute on the input tree, but there will be many DIEs with that attribute pointing to the same file entry (same index into the CU's .debug_ling table, so getting the same dwarf::source_file again and again). Hence, we expect to see the actual same pointer (in a dwarf reader object, really the same pointer into the mmapped DWARF file data) appearing many times in the input tree's attr_values. Matching pointers you've seen before is quicker than matching string values you've seen before. So the copier keeps maps keyed by these string (and file entry) pointers into the input data, to find exact attr_value's already found/created to match other instances of the same input string (or file entry). (This sped things up a fair bit when I added it, though I don't recall the details.) For all the simple attribute values the method is simple. For each flavor of attr_value, the collector holds a set of values seen before. The subr::value_set container type is used for this. It's based on the STL unordered_set, which is a hash table where we supply the key type and the hash and comparison functions on that type. In these sets, our key types are dwarf_output::value::value_*, which are the flavor-specific subtypes that an attr_value actually points to. So, each of those types has to provide an integer hash function (a "hasher" inner class) and a comparison function (operator==). The copier does a set insertion of the input value object, which installs a new dwarf_output::value::value_foobar object if none matched, or doesn't if one did, and then produces an attr_value that points to that. Note that since this is the only way any dwarf_output::attr_value objects can ever get created, within one dwarf_output object (really, one collector), simple pointer equality is exactly correct as a semantic equality check for the non-reference attribute values. A corollary is that just this pointer value itself can be used as the hash value for the attr_value object (a perfect hash function). If there are no reference attributes to consider, then it proceeds from the bottom up in the same straightforward way. That is, the entire map of attributes (considered as a sequence of <int name, attr_value> pairs in a canonical name-keyed order) can be hashed together to get a hash value for the attributes container. Then a leaf node can combine the tag into the that hash value to get a hash value for the whole DIE. Then, a sequence of sibling leaf nodes' hash values are combined to form a hash value for the whole children container. Then that hash value is combined with the tag and the attribute map hash value to form a hash value for the whole DIE. Then on up. So this is what we do. The simple attribute values are hashed on the actual values (string data, integer values, etc.), and then those pointer values along with the attribute name (integers) in a canonical order are combined with the DIE tag into a single integer hash. The _m_attr_sets set (i.e. hash table) holds these <attribute map, tag integer> pairs. Then children vectors, starting with the empty vector for leaf DIEs, are hashed into the _m_broods set. (The empty vector has hash value zero.) Finally, a dwarf_output::debug_info_entry just consists of pointers to the attributes map and children vector, and the DIE tag integer. We just combine those two pointer values and the tag to make the integer hash of the debug_info_entry object itself. _m_unique holds the set of unique debug_info_entry objects. So, a leaf DIE gets created with an attributes map pointer and a pointer to the one and only empty children vector object. Then a non-empty children vector is simply a vector of pointers to debug_info_entry objects like that one. Like for the other sets, a pointer to a particular debug_info_entry in the collector serves as a perfect integer hash, since by definition pointer equality of the debug_info_entry object exactly maps to semantic equivalence. So, the children vector is a simple container of pointers, and the vector itself has a hash value made just by combining the pointer values as so many constituent integer hash values. So, it's like that when no attribute values are references. The exact control flow between the constructors of the dwarf_output inner classes and dwarf_output::copier is extremely hairy. But the extreme template weirdness is mostly just to get the copier pointer passed down the levels of constructors to where it calls an add_* method on the copier. It's there, and the collector methods they call, where the real work happens. It's even like that when there are some references, for simple cases. So, start with a completely simple leaf node: <base_type name="int" byte_size=0x4 encoding=signed/> That's all simple, so its attribute values, map, empty children vector, debug_info_entry will all be hashed and inserted in the collector as above. Now there is a known debug_info_entry pointer for that DIE. Next, say you are copying: <variable name="x" type="#int" location=.../> The various simple attributes all resolve to known value pointers. The type="#int" attribute is the only reference value, pointing to the DIE we just mentioned. Since that's a fully resolved known DIE now, we can in fact treat it the same way as we treat simple attribute values. Along with the one debug_info_entry object in the collector, we only need to have exactly one dwarf_output::value::value_reference object that points to it. So here, that value object pointer is perfect hash and equality test for a reference attribute that refers to a DIE identical to the "int" above, just like the value object pointers for the simple value flavors are. If that's all there were to do, it would be easy. Of course, there are circular references. So say we have: <structure_type id="t1" name="list"> <member name="next" type="#p1"/> </structure_type> <pointer_type id="p1" type="#t1"/> Now, the structure_type's own attribute map is simple, so that hashes up nicely. But, the member's attribute map includes a circular reference. So, what we do is still hash up the whole attribute map, but use zero as the hash value for the reference value. Now <member/> as a whole has a hash value that contributes to the children vector hash value, and that contributes to the structure_type's overall hash value. Two entries that look like this but have different references in the type attribute--but both circular references in the overall graph--would have the same hash value. So when comparing those, we'd get a hash collision and have to do a full comparison. That would find the two attribute maps identical, so look at the two children vectors. The corresponding member entries' attribute maps would then be compared for equivalence, the non-reference attribute values (i.e. name="next") matching by simple pointer comparison. But for the reference, we have to do a ref-walking comparison in the manner of dwarf_comparator. This plan means that for <pointer_type type="something circular"/> and <reference_type type="something circular"/>, we'll actually have dozens to thousands of entries with the same hash value--basically every single pointer_type to a nontrivial class/struct is a hash collision with every other. If that's the plan, then the ref-walking comparisons we do have to be very well-optimized. We'll be doing many, many of them every time we come across another pointer_type entry in the input. Picayune as I've been, this is just the broad overview. It seems solid, and the obvious thing to do, up to the point of the plan for circular references. It's not entirely clear that having so many hash collisions in this approach is a reasonable way to go. But I don't know what a different plan would be for efficiently identifying equivalent subtrees in the presence of circular references. Most of the real complexity in the copier has to do with handling forward references in attributes, including circular references. Aside from the simple-value mapping caches I described earlier, everything else is pretty much just interface glue for making the dwarf_output constructors work. The possibility of forward references means we have to partially construct entries while waiting to resolve their references, so there are lots of internal data structures for that. In valid input data, a single entire (logical) CU tree has no dangling references. So all the partially-built tracking data structures only need to live (at most) during the construction of one compile_unit tree in the dwarf_output object. The only things that live longer in the copier are just caches keyed on parts of the input that might be shared across input logical CUs (like the strings, but also input DIEs--for when eventually reading compressed format input). For following the control flow, you can start at dwarf_output::copier::make_unit. You can read the classes it calls to see how it's all organized. (Note how at various levels, most of the actual work happens inside a foo_copier constructor.) For all of the entry-copying hair inside dwarf_output::copier, I think it will be easiest to just answer questions as you read through the code, rather than trying to describe a structure and a plan. I can explain it all piece by piece, but at this point will largely be recovering the knowledge from re-reading the code anyway, so we can read it together. How that all works is quite ad hoc (and about the third generation of how I organized it), and I probably couldn't articulate an overall algorithm it is implementing. There is: #if 0 // Toggle this to enable massive debugging spew during construction. and making that 1 will get you quite a lot of spew that may help get a handle on how the partial entry resolution control flow goes. The usual ways to drive a test of this is: tests/dwarf-print --output file For nontrivial CUs in the input file, the debugging spew is truly massive. Thanks, Roland

2 30

g++-4.5 vs dwarf branch
by Mark Wielaard 30 Nov '10

30 Nov '10

Hi, I happily updated my machine to fedora 14 which comes with g++ 4.5.1. Unfortunately that g++ version doesn't like the dwarf branch libdw/c++ stuff. There seem to be three issues. I solved one and put that on the mjw/gcc45 branch. Reviews to see if it could be moved to the dwarf branch would be nice. - g++ 4.5 doesn't like to call constructors by name directly. This makes it impractical to have a class and a function that tries to call the constructor by the same name. So rename the various compile_units classes to compile_units_type. Leaving the function compile_units to return the correct instance type. (This is on the branch.) - std::pair<_T1, _T2>::second’ has incomplete type. This is strange. I haven't been able to reproduce it in a smaller setup. The error looks as follows: /usr/lib/gcc/x86_64-redhat-linux/4.5.1/../../../../include/c ++/4.5.1/bits/stl_pair.h: In instantiation of ‘std::pair<const elfutils::dwarf_output::debug_info_entry, elfutils::dwarf_output::die_info>’: /home/mark/src/elfutils/libdw/c++/dwarf_output:233:63: instantiated from here /usr/lib/gcc/x86_64-redhat-linux/4.5.1/../../../../include/c ++/4.5.1/bits/stl_pair.h:77:11: error: ‘std::pair<_T1, _T2>::second’ has incomplete type /home/mark/src/elfutils/libdw/c++/dwarf_output:146:12: error: forward declaration of ‘struct elfutils::dwarf_output::die_info’ Which doesn't make sense to me. The code it complains about is an invocation of vector<die_info_pair *>.end(). The following "patch" makes it go away. But I have no idea why, both code constructs seem to not actually use the pair directly: diff --git a/libdw/c++/dwarf_output b/libdw/c++/dwarf_output index 9e5c1f6..3a22b7a 100644 --- a/libdw/c++/dwarf_output +++ b/libdw/c++/dwarf_output @@ -230,8 +230,10 @@ namespace elfutils inline void set_hash () { _m_hash = 0; +/* for (_base::iterator i = _base::begin (); i != _base::end (); ++i) subr::hash_combine (_m_hash, (uintptr_t) *i); +*/ } inline children_type () {} @@ -411,7 +413,7 @@ namespace elfutils for the circular_reference case, below. */ virtual size_t hash () const { - return ref->identity (); + return 0; // ref->identity (); } }; - Last issue is the "swap" function from compile_units[_type]: libdw/c++/dwarf_output:566:7: error: no matching function for call to ‘elfutils::dwarf_output::compile_units_type::swap(elfutils::dwarf_output::compile_units_type)’ I am not clear where the swap function should have been defined. Cheers, Mark

2 5

elf utils manual?
by Fennec Cleric 30 Nov '10

30 Nov '10

Hello all, I'm a college student working on a final project for my computer security class. I am trying to manipulate .ELF files and I found this elf-utils.rpm that looks like it will be extremely useful in reading, manipulating, and writing a .ELF file. However, there is no man file and I haven't been able to find any API files or documentation online. Could someone point me to some online documentation or explain how to use elf utils? Thanks, FF

3 2

Re: 0.150 release now?
by Roland McGrath 22 Nov '10

22 Nov '10

I committed NEWS entries, and my test builds are all OK (Fedora 13/14/rawhide, RHEL 5/6). So I think 0.150 is ready to sail. Thanks, Roland

1 1

0.150 release now?
by Roland McGrath 22 Nov '10

22 Nov '10

I'd like to have an 0.150 release real soon now. The main reason is the libdwfl-vs-prelink fix (1743d7f). The libdw-huge-alloca fix (5f95e2b) is the only other really notable thing since 0.149. But since that libdwfl bug is holding back systemtap users, I'd like to make a release without much more delay. Anybody have any issues with that? I haven't done test rpm builds yet, but I will do those this evening and report here when I think we're ready enough to go. Ulrich, if you can turn the crank quickly (e.g. Tuesday or Wednesday) then that would be great. If not and you don't object, then I may just do it all myself, probably Wednesday. For those of us on the US holiday schedule, this is a short week, so what doesn't happen by Wednesday won't happen until at least Monday, which is probably fine enough too. Thanks, Roland

1 0

dwarf_edit_output testcase sanity checking
by Mark Wielaard 17 Nov '10

17 Nov '10

Hi, On the dwarf branch I added a bunch of "simple" small dwarf_edit examples to tests/dwarf_edit_output.cc. The testcases are then used to compare some DIEs with dwarf_comparator, to create a dwarf_output object, check whether expected DIEs have been merged and whether the dwarf_edit and dwarf_output objects are still comparable. If you #define _DWARF_OUTPUT_DEBUG_SPEW 1 in dwarf_output it is a nice way to follow the dwarf_output copier logic. There is one testcase for which dwarf_comparator and dwarf_output don't agree on whether they are equal/sharable. Lets run it with: $ tests/dwarf_edit_output 10 inout (Runs test number 10, shows both input, the dwarf_edit object, as output, the dwarf_output object. run with --offsets to see the raw offsets.) *two_circular_structs2* dwarf_edit: <compile_unit producer="dwarf_edit_output_test"> <base_type ref="ref2" name="int" byte_size=0x4/> <pointer_type ref="ref3" byte_size=0x8 type="#ref1"/> <structure_type ref="ref1" name="list" byte_size=0x10> <member name="i" type="#ref2"/> <member name="next" type="#ref3"/> </structure_type> <variable name="var" type="#ref3"/> </compile_unit> <compile_unit producer="dwarf_edit_output_test"> <base_type ref="ref4" name="int" byte_size=0x4/> <structure_type ref="ref6" name="list" byte_size=0x10> <member name="i" type="#ref4"/> <member name="next" type="#ref5"/> </structure_type> <pointer_type ref="ref5" byte_size=0x8 type="#ref6"/> <variable name="var" type="#ref5"/> </compile_unit> [At this point there will be a check that both vars in the two CUs dwarf_comparator equal, and they do. But...] dwarf_output: <compile_unit producer="dwarf_edit_output_test"> <base_type ref="ref8" name="int" byte_size=0x4/> <pointer_type ref="ref9" byte_size=0x8 type="#ref7"/> <structure_type ref="ref7" name="list" byte_size=0x10> <member name="i" type="#ref8"/> <member name="next" type="#ref9"/> </structure_type> <variable name="var" type="#ref9"/> </compile_unit> <compile_unit producer="dwarf_edit_output_test"> <base_type ref="ref10" name="int" byte_size=0x4/> <structure_type ref="ref12" name="list" byte_size=0x10> <member name="i" type="#ref10"/> <member name="next" type="#ref11"/> </structure_type> <pointer_type ref="ref11" byte_size=0x8 type="#ref12"/> <variable name="var" type="#ref11"/> </compile_unit> offset last var cu1: 0x871800 offset last var cu2: 0x873860 tests/dwarf_edit_output: fail test_last_two_var_dies #10 'two_circular_structs2' So dwarf_output doesn't think so. Now logically they seem to be the same. So I think dwarf_comparator is correct. But a case could be made that although they are semantically the same, they aren't really structurally equal. In this context the order of the DIEs doesn't matter (top-level CU children). But in theory the order of children DIEs does matter in some contexts. So, is the above testcase something where we really want to mark both vars as equal? Thanks, Mark

3 17

RFC: Cross-compiling elfutils
by rmorell＠nvidia.com 16 Nov '10

16 Nov '10

Hi, I'm building elfutils on x86_64 (the "build machine", using the new autotools terminology) for ARM (the "host machine"). Unfortunately, this fails because the build system emits an ARM binary i386_gendis, and then attempts to execute this binary on x86_64. I have a patch to fix this, which defines a FOR_BUILD compiler that's distinct from the default compiler when cross-compiling, and uses that compiler to build i386_gendis. I can provide this patch. However, there is one more tricky part: Unfortunately, simply using this compiler to build the i386_gendis sources isn't sufficient; i386_gendis also expects to have libeu.a linked in, which is built once and then linked into multiple binaries. It turns out that the only thing i386_gendis needs libeu.a for is the xmalloc wrapper and friends, which provide a helpful call to error() when the call fails. For my purposes, I just replaced the xmalloc() calls to malloc() to remove i386_gendis's dependence on libeu.a. This seems fairly safe, given that unlike general uses of elfutils this particular usage is not unbounded and is unlikely to fail memory allocations. This does seem kind of lame though; I guess an alternative would be to build libeu.a for both build and host arch when cross-compiling. Would removing libeu.a from i386_gendis be acceptable? Thanks, Robert ----------------------------------------------------------------------------------- This email message is for the sole use of the intended recipient(s) and may contain confidential information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. -----------------------------------------------------------------------------------

1 0

DSO: library or PIE
by Kirill A. Shutemov 15 Nov '10

15 Nov '10

Is there any easy way to detect whether DSO is PIE? I only know about tricky heuristic: http://git.altlinux.org/people/ldv/packages/rpm.git?p=rpm.git;a=blob;f=scri… -- Kirill A. Shutemov

2 3

[PATCH] Fix findtextrel to use DT_SYMTAB when present
by Mike Hommey 11 Nov '10

11 Nov '10

findtextrel skips going through all sections when it encounters the SHT_DYNAMIC section, and that section indicates text relocations through DT_TEXTREL or DF_TEXTREL. But in most cases, the DT_SYMTAB section appears after the SHT_DYNAMIC section, in which case findtextrel never takes advantage of that section. This fixes it. --- src/findtextrel.c | 10 ++++++---- 1 files changed, 6 insertions(+), 4 deletions(-) diff --git a/src/findtextrel.c b/src/findtextrel.c index 9d10982..88276c2 100644 --- a/src/findtextrel.c +++ b/src/findtextrel.c @@ -207,6 +207,7 @@ noop (void *arg __attribute__ ((unused))) static int process_file (const char *fname, bool more_than_one) { + int have_textrel = 0; int result = 0; void *knownsrcs = NULL; @@ -297,17 +298,18 @@ process_file (const char *fname, bool more_than_one) if (dyn->d_tag == DT_TEXTREL || (dyn->d_tag == DT_FLAGS && (dyn->d_un.d_val & DF_TEXTREL) != 0)) - goto have_textrel; + have_textrel = 1; } } else if (shdr->sh_type == SHT_SYMTAB) symscn = scn; } - error (0, 0, gettext ("no text relocations reported in '%s'"), fname); - return 1; + if (!have_textrel) { + error (0, 0, gettext ("no text relocations reported in '%s'"), fname); + return 1; + } - have_textrel:; int fd2 = -1; Elf *elf2 = NULL; /* Get the address ranges for the loaded segments. */ -- 1.7.2.3

2 3

i386 return value
by Kurt Roeckx 08 Nov '10

08 Nov '10

Hi, In i386_retal.c we have: /* %eax, or pair %eax, %edx. */ static const Dwarf_Op loc_intreg[] = { { .atom = DW_OP_reg0 }, { .atom = DW_OP_piece, .number = 4 }, { .atom = DW_OP_reg1 }, { .atom = DW_OP_piece, .number = 4 }, }; In i386_regs.c we have: switch (regno) { static const char baseregs[][2] = { "ax", "cx", "dx", "bx", "sp", "bp", "si", "di", "ip" }; It seems to comment doesn't match the code. As far as I can see the order of the registers in i386_regs.c is correct, and different then x86_64. Looking at the gdb source, edx is used for the return value. So I believe the attached patch should fix things. It still passes the regression tests after the change. The issue was pointed out by Thorsten Glaser <tg(a)mirbsd.de> Kurt

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

elfutils-devel November 2010