Handle most of type changes, e.g. const is changed to struct,
or struct to pointers. In all these cases we create default args.
They may not give the coverage anymore, but still better than
losing them right away.
We now always create net namespace for testing,
so socket ports and other IDs do not overlap between
different test processes.
Proc types play badly with squashing packets to ANYBLOB.
To squash into a block we need concrete value, but it depends
on process id.
Removing proc also makes tun setup and address descriptions simpler.
Squash complex structs into flat byte array and mutate this array
with generic blob mutations. This allows to mutate what we currently
consider as paddings and add/remove paddings from structs, etc.
1. mmap all memory always, without explicit mmap calls in the program.
This makes lots of things much easier and removes lots of code.
Makes mmap not a special syscall and allows to fuzz without mmap enabled.
2. Change address assignment algorithm.
Current algorithm allocates unmapped addresses too frequently
and allows collisions between arguments of a single syscall.
The new algorithm analyzes actual allocations in the program
and places new arguments at unused locations.
This reduces size of a corpus in half.
We store corpus on manager and on hub,
so this will reduce their memory consumption.
But also makes large programs more readable.
Introduce isUsed(arg) helper, use it in several places.
Move method definitions closer to their types.
Simplify presence check for ArgUsed.Used() in several places.
Fixes#188
We now will write just ""/1000 to denote a 1000-byte output buffer.
Also we now don't store 1000-byte buffer in memory just to denote size.
Old format is still parsed.
Fixes#460
File names, crypto algorithm names, etc in programs are completely unreadable:
bind$alg(r0, &(0x7f0000408000)={0x26, "6861736800000000000000000000",
0x0, 0x0, "6d6435000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000000000000000000
00000000000"}, 0x58)
Introduce another format for printable strings.
New args are denoted by '' ("" for old args).
New format is enabled for printable chars, \x00
and \t, \r, \n.
Example:
`serialize(&(0x7f0000408000)={"6861736800000000000000000000", "4849000000"})`,
vs:
`serialize(&(0x7f0000408000)={'hash\x00', 'HI\x00'})`,
For string[N] we successfully deserialize a string of any length.
Similarly for a fixed-size array[T, N] we successfully deserialize
an array of any size.
Such programs later crash in foreachSubargOffset because static size
Type.Size() does not match what we've calculated iterating over fields.
The crash happens only in SerializeForExec in syz-fuzzer,
which is especially bad.
Fix this from both sides:
1. Validate sizes of arrays/buffers in Validate.
2. Repair incorrect sizes in Deserialize.
Now each prog function accepts the desired target explicitly.
No global, implicit state involved.
This is much cleaner and allows cross-OS/arch testing, etc.
Large overhaul moves syscalls and arg types from sys to prog.
Sys package now depends on prog and contains only generated
descriptions of syscalls.
Introduce prog.Target type that encapsulates all targer properties,
like syscall list, ptr/page size, etc. Also moves OS-dependent pieces
like mmap call generation from prog to sys.
Update #191
We currently use uintptr for all values.
This won't work for 32-bit archs.
Moreover in some cases we use uintptr but assume
that it is always 64-bits (e.g. in encodingexec).
Switch everything to uint64.
Update #324
After a change in syscall description the number of syscall arguments
might change and some of the programs in corpus get invalidated.
This change makes syzkaller to generate missing arguments when decoding a
program as an attempt to fix and keep more programs from corpus.
Right now Arg is a huge struct (160 bytes), which has many different fields
used for different arg kinds. Since most of the args we see in a typical
corpus are ArgConst, this results in a significant memory overuse.
This change:
- makes Arg an interface instead of a struct
- adds a SomethingArg struct for each arg kind we have
- converts all *Arg pointers into just Arg, since interface variable by
itself contains a pointer to the actual data
- removes ArgPageSize, now ConstArg is used instead
- consolidates correspondence between arg kinds and types, see comments
before each SomethingArg struct definition
- now LenType args that denote the length of VmaType args are serialized as
"0x1000" instead of "(0x1000)"; to preserve backwards compatibility
syzkaller is able to parse the old format for now
- multiple small changes all over to make the above work
After this change syzkaller uses twice less memory after deserializing a
typical corpus.
The optimization change removed validation too aggressively.
We do need program validation during deserialization,
because we can get bad programs from corpus or hub.
Restore program validation after deserialization.
A bunch of spot optmizations after cpu/memory profiling:
1. Optimize hot-path coverage comparison in fuzzer.
2. Don't allocate and copy serialized program, serialize directly into shmem.
3. Reduce allocations during parsing of output shmem (encoding/binary sucks).
4. Don't allocate and copy coverage arrays, refer directly to the shmem region
(we are not going to mutate them).
5. Don't validate programs outside of tests, validation allocates tons of memory.
6. Replace the choose primitive with simpler switches.
Choose allocates fullload of memory (for int, func, and everything the func refers).
7. Other minor optimizations.
Currently we generate arrays of size [0,5] with equal probability.
Generate [0,10] with bias towards smaller arrays. But 0 has the lowest probability.
I've benchmark a slightly different change with max array size of 20,
results are somewhat inconclusive: it was better than baseline almost all way,
but baseline suddenly caught up at the end. It also considerably reduced
executions per second (by ~20%). So increasing array size to 10 should be a win...
bufio.Scanner has a default limit of 4K per line,
if a program contains longer line, it fails.
Extend the limit to 64K.
Also check scanning errors. Turns out even scanning of bytes.Buffer
can fail due to the line limit.
Eliminate assignTypeAndDir function and instead assign
types to all args during construction.
This will allow considerable simplifation of assignSizes.
Currently we store most types by value in sys.Type.
This is somewhat counter-intuitive for C++ programmers,
because one can't easily update the type object.
Store pointers to type objects for all types.
It also makes it easier to update types, e.g. adding paddings.
