[1/2] Add unit test to aarch64 prologue analyzer
Commit Message
We don't have an effective way to test prologue analyzer which is
highly dependent on instruction patterns in prologue generated by
compiler. GDB prologue analyzer may not handle the new sequences
generated by new compiler, or may still handle some sequences that
generated by very old compilers which are no longer used. The
former is a functionality issue, while the latter is a maintenance
issue.
The input and output of prologue analyzer is quite clear, so it
fits for unit test. The input is series of instructions, and the
output are 1) where prologue end, 2) where registers are saved.
In aarch64, they are represented in 'struct aarch64_prologue_cache'.
This patch refactors aarch64_analyze_prologue so it can read
instructions from either real target or test harness. In unit
test aarch64_analyze_prologue_test, aarch64_analyze_prologue gets
instructions we prepared in the test, as the input of prologue
analyzer. Then, we checked various fields in
'struct aarch64_prologue_cache'.
gdb:
2016-11-28 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c: Include "selftest.h".
(abstract_instruction_reader): New class.
(instruction_reader): New class.
(aarch64_analyze_prologue): Add new parameter reader. Call
reader.read instead of read_memory_unsigned_integer.
[GDB_SELF_TEST] (instruction_reader_test): New class.
(aarch64_analyze_prologue_test): New function.
(_initialize_aarch64_tdep) [GDB_SELF_TEST]: Register
selftests::aarch64_analyze_prologue_test.
* trad-frame.c (trad_frame_cache_zalloc):
(trad_frame_alloc_saved_regs): Add a new function.
* trad-frame.h (trad_frame_alloc_saved_regs): Declare.
---
gdb/aarch64-tdep.c | 116 ++++++++++++++++++++++++++++++++++++++++++++++++++++-
gdb/trad-frame.c | 21 ++++++----
gdb/trad-frame.h | 1 +
3 files changed, 129 insertions(+), 9 deletions(-)
Comments
Yao Qi writes:
> @@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> return start;
> }
>
> +static CORE_ADDR
> +aarch64_analyze_prologue (struct gdbarch *gdbarch,
> + CORE_ADDR start, CORE_ADDR limit,
> + struct aarch64_prologue_cache *cache)
> +{
> + instruction_reader reader { };
> +
Could we use the default constructor here? If it's kept.
> + return aarch64_analyze_prologue (gdbarch, start, limit, cache,
> + reader);
> +}
> +
> +#if GDB_SELF_TEST
> +
> +namespace selftests {
> +
> + /* Instruction reader from manually cooked instruction sequences. */
> + class instruction_reader_test : public abstract_instruction_reader
> + {
> + public:
> + instruction_reader_test() = default ;
Very nit, but there's a space before ';'
Also I wonder if we need to specify the default constructor explicitly ?
Is there a rationale for it?
It's never used too, unless you apply my previous comment.
On Tue, Nov 29, 2016 at 09:57:46AM -0500, Antoine Tremblay wrote:
>
> Yao Qi writes:
>
> > @@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> > return start;
> > }
> >
> > +static CORE_ADDR
> > +aarch64_analyze_prologue (struct gdbarch *gdbarch,
> > + CORE_ADDR start, CORE_ADDR limit,
> > + struct aarch64_prologue_cache *cache)
> > +{
> > + instruction_reader reader { };
> > +
>
> Could we use the default constructor here? If it's kept.
>
It uses the default constructor. instruction_reader has the default
constructor.
+ public:
+ instruction_reader () = default;
Do you mean I need to remove it?
> > + return aarch64_analyze_prologue (gdbarch, start, limit, cache,
> > + reader);
> > +}
> > +
> > +#if GDB_SELF_TEST
> > +
> > +namespace selftests {
> > +
> > + /* Instruction reader from manually cooked instruction sequences. */
> > + class instruction_reader_test : public abstract_instruction_reader
> > + {
> > + public:
> > + instruction_reader_test() = default ;
>
> Very nit, but there's a space before ';'
>
Oh, yes, it should be "instruction_reader_test () = default;"
> Also I wonder if we need to specify the default constructor explicitly ?
> Is there a rationale for it?
>
> It's never used too, unless you apply my previous comment.
The instruction_reader_test default constructor is never used. How
about using "= delete"?
instruction_reader_test () = delete;
instruction_reader_test (std::initializer_list<uint32_t> init)
: insns{init} {}
Yao Qi writes:
> On Tue, Nov 29, 2016 at 09:57:46AM -0500, Antoine Tremblay wrote:
>>
>> Yao Qi writes:
>>
>> > @@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
>> > return start;
>> > }
>> >
>> > +static CORE_ADDR
>> > +aarch64_analyze_prologue (struct gdbarch *gdbarch,
>> > + CORE_ADDR start, CORE_ADDR limit,
>> > + struct aarch64_prologue_cache *cache)
>> > +{
>> > + instruction_reader reader { };
>> > +
>>
>> Could we use the default constructor here? If it's kept.
>>
>
> It uses the default constructor. instruction_reader has the default
> constructor.
>
> + public:
> + instruction_reader () = default;
>
> Do you mean I need to remove it?
>
No, sorry I need to improve on C++11 I thought you needed a
std::initializer_list to init with {} like in the _test class but it
it works with the default constructor.
>> > + return aarch64_analyze_prologue (gdbarch, start, limit, cache,
>> > + reader);
>> > +}
>> > +
>> > +#if GDB_SELF_TEST
>> > +
>> > +namespace selftests {
>> > +
>> > + /* Instruction reader from manually cooked instruction sequences. */
>> > + class instruction_reader_test : public abstract_instruction_reader
>> > + {
>> > + public:
>> > + instruction_reader_test() = default ;
>>
>> Very nit, but there's a space before ';'
>>
>
> Oh, yes, it should be "instruction_reader_test () = default;"
>
>> Also I wonder if we need to specify the default constructor explicitly ?
>> Is there a rationale for it?
>>
>> It's never used too, unless you apply my previous comment.
>
> The instruction_reader_test default constructor is never used. How
> about using "= delete"?
>
> instruction_reader_test () = delete;
> instruction_reader_test (std::initializer_list<uint32_t> init)
> : insns{init} {}
Yes that would be more appropriate if we're going to specify that.
I just wrote a patch with a C++ class and did not include explicit
default constructors do you think we should make it a code convention to
explicitly specify their existence or non-existence (=default, =delete) ?
I could not find mention of that in GCC's C++ conventions...
On 11/29/2016 02:12 PM, Yao Qi wrote:
> We don't have an effective way to test prologue analyzer which is
> highly dependent on instruction patterns in prologue generated by
> compiler. GDB prologue analyzer may not handle the new sequences
> generated by new compiler, or may still handle some sequences that
> generated by very old compilers which are no longer used. The
> former is a functionality issue, while the latter is a maintenance
> issue.
>
> The input and output of prologue analyzer is quite clear, so it
> fits for unit test. The input is series of instructions, and the
> output are 1) where prologue end, 2) where registers are saved.
> In aarch64, they are represented in 'struct aarch64_prologue_cache'.
>
> This patch refactors aarch64_analyze_prologue so it can read
> instructions from either real target or test harness. In unit
> test aarch64_analyze_prologue_test, aarch64_analyze_prologue gets
> instructions we prepared in the test, as the input of prologue
> analyzer. Then, we checked various fields in
> 'struct aarch64_prologue_cache'.
>
> gdb:
>
> 2016-11-28 Yao Qi <yao.qi@linaro.org>
>
> * aarch64-tdep.c: Include "selftest.h".
> (abstract_instruction_reader): New class.
> (instruction_reader): New class.
> (aarch64_analyze_prologue): Add new parameter reader. Call
> reader.read instead of read_memory_unsigned_integer.
> [GDB_SELF_TEST] (instruction_reader_test): New class.
> (aarch64_analyze_prologue_test): New function.
> (_initialize_aarch64_tdep) [GDB_SELF_TEST]: Register
> selftests::aarch64_analyze_prologue_test.
> * trad-frame.c (trad_frame_cache_zalloc):
> (trad_frame_alloc_saved_regs): Add a new function.
> * trad-frame.h (trad_frame_alloc_saved_regs): Declare.
> ---
> gdb/aarch64-tdep.c | 116 ++++++++++++++++++++++++++++++++++++++++++++++++++++-
> gdb/trad-frame.c | 21 ++++++----
> gdb/trad-frame.h | 1 +
> 3 files changed, 129 insertions(+), 9 deletions(-)
>
> diff --git a/gdb/aarch64-tdep.c b/gdb/aarch64-tdep.c
> index 6b95d7c..b10002a 100644
> --- a/gdb/aarch64-tdep.c
> +++ b/gdb/aarch64-tdep.c
> @@ -44,6 +44,7 @@
> #include "infcall.h"
> #include "ax.h"
> #include "ax-gdb.h"
> +#include "selftest.h"
>
> #include "aarch64-tdep.h"
>
> @@ -195,6 +196,29 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
> fprintf_filtered (file, _("AArch64 debugging is %s.\n"), value);
> }
>
> +/* Abstract instruction reader. */
> +
> +class abstract_instruction_reader
> +{
> +public:
> + /* Read in one instruction. */
> + virtual ULONGEST read (CORE_ADDR memaddr, int len,
> + enum bfd_endian byte_order) = 0;
> +};
> +
> +/* Instruction reader from real target. */
> +
> +class instruction_reader : public abstract_instruction_reader
> +{
> + public:
> + instruction_reader () = default;
(As mentioned in the previous email, this just looks like
unnecessary redundancy to me; suggest just removing it. The compiler
generates it for you.)
> +
> + ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
> + {
> + return read_memory_unsigned_integer (memaddr, len, byte_order);
> + }
> +};
> +
> /* Analyze a prologue, looking for a recognizable stack frame
> and frame pointer. Scan until we encounter a store that could
> clobber the stack frame unexpectedly, or an unknown instruction. */
> @@ -202,7 +226,8 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
> static CORE_ADDR
> aarch64_analyze_prologue (struct gdbarch *gdbarch,
> CORE_ADDR start, CORE_ADDR limit,
> - struct aarch64_prologue_cache *cache)
> + struct aarch64_prologue_cache *cache,
> + abstract_instruction_reader& reader)
> {
> enum bfd_endian byte_order_for_code = gdbarch_byte_order_for_code (gdbarch);
> int i;
> @@ -221,7 +246,7 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> uint32_t insn;
> aarch64_inst inst;
>
> - insn = read_memory_unsigned_integer (start, 4, byte_order_for_code);
> + insn = reader.read (start, 4, byte_order_for_code);
>
> if (aarch64_decode_insn (insn, &inst, 1) != 0)
> break;
> @@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> return start;
> }
>
> +static CORE_ADDR
> +aarch64_analyze_prologue (struct gdbarch *gdbarch,
> + CORE_ADDR start, CORE_ADDR limit,
> + struct aarch64_prologue_cache *cache)
> +{
> + instruction_reader reader { };
It's more idiomatic to just do:
instruction_reader reader;
when you want default construction.
> +
> + return aarch64_analyze_prologue (gdbarch, start, limit, cache,
> + reader);
> +}
> +
> +#if GDB_SELF_TEST
> +
> +namespace selftests {
> +
> + /* Instruction reader from manually cooked instruction sequences. */
> + class instruction_reader_test : public abstract_instruction_reader
> + {
> + public:
> + instruction_reader_test() = default ;
> + instruction_reader_test (std::initializer_list<uint32_t> init)
> + : insns{init} {}
I think we should put a space before "{" -> "insns {init}". We put
it before "(", and before "{" in all other contexts.
> +
> + ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
> + {
> + SELF_CHECK (len == 4);
> + SELF_CHECK (memaddr % 4 == 0);
> + SELF_CHECK (memaddr / 4 < insns.size());
> +
> + return insns[memaddr / 4];
> + }
> +
> + private:
> + std::vector<uint32_t> insns;
Private data members should be prefixed with "m_".
I'll note that it always itches me a bit when we do
unnecessary copying. :-) In this case, you always
start from an array of instructions known at compile-time,
and copy it into the vector at run time. You could
instead create the instructions array as a separate const
array, and pass than to the reader's constructor
as parameter, which would store a pointer to the array,
instead of a deep copy.
> + };
> +
> +static void
> +aarch64_analyze_prologue_test (void)
> +{
> + struct gdbarch_info info;
> +
> + gdbarch_info_init (&info);
> + info.bfd_arch_info = bfd_scan_arch ("aarch64");
> +
> + struct gdbarch *gdbarch = gdbarch_find_by_info (info);
> + SELF_CHECK (gdbarch != NULL);
> +
> + struct aarch64_prologue_cache cache;
> + cache.saved_regs = trad_frame_alloc_saved_regs (gdbarch);
> +
> + instruction_reader_test test {
> + 0xa9af7bfd, /* stp x29, x30, [sp,#-272]! */
> + 0x910003fd, /* mov x29, sp */
> + 0x97ffffe6, /* bl 0x400580 */
> + };
Indentation looks odd here. "0x..." should be two columns
to the right of instruction_reader_test, and "};" aligned
at the same level as "instruction_reader_test".
Thanks,
Pedro Alves
On 11/29/2016 08:12 AM, Yao Qi wrote:
> We don't have an effective way to test prologue analyzer which is
> highly dependent on instruction patterns in prologue generated by
> compiler. GDB prologue analyzer may not handle the new sequences
> generated by new compiler, or may still handle some sequences that
> generated by very old compilers which are no longer used. The
> former is a functionality issue, while the latter is a maintenance
> issue.
>
> The input and output of prologue analyzer is quite clear, so it
> fits for unit test. The input is series of instructions, and the
> output are 1) where prologue end, 2) where registers are saved.
> In aarch64, they are represented in 'struct aarch64_prologue_cache'.
>
> This patch refactors aarch64_analyze_prologue so it can read
> instructions from either real target or test harness. In unit
> test aarch64_analyze_prologue_test, aarch64_analyze_prologue gets
> instructions we prepared in the test, as the input of prologue
> analyzer. Then, we checked various fields in
> 'struct aarch64_prologue_cache'.
>
> gdb:
>
> 2016-11-28 Yao Qi <yao.qi@linaro.org>
>
> * aarch64-tdep.c: Include "selftest.h".
> (abstract_instruction_reader): New class.
> (instruction_reader): New class.
> (aarch64_analyze_prologue): Add new parameter reader. Call
> reader.read instead of read_memory_unsigned_integer.
> [GDB_SELF_TEST] (instruction_reader_test): New class.
> (aarch64_analyze_prologue_test): New function.
> (_initialize_aarch64_tdep) [GDB_SELF_TEST]: Register
> selftests::aarch64_analyze_prologue_test.
> * trad-frame.c (trad_frame_cache_zalloc):
> (trad_frame_alloc_saved_regs): Add a new function.
> * trad-frame.h (trad_frame_alloc_saved_regs): Declare.
> ---
> gdb/aarch64-tdep.c | 116 ++++++++++++++++++++++++++++++++++++++++++++++++++++-
> gdb/trad-frame.c | 21 ++++++----
> gdb/trad-frame.h | 1 +
> 3 files changed, 129 insertions(+), 9 deletions(-)
>
> diff --git a/gdb/aarch64-tdep.c b/gdb/aarch64-tdep.c
> index 6b95d7c..b10002a 100644
> --- a/gdb/aarch64-tdep.c
> +++ b/gdb/aarch64-tdep.c
> @@ -44,6 +44,7 @@
> #include "infcall.h"
> #include "ax.h"
> #include "ax-gdb.h"
> +#include "selftest.h"
>
> #include "aarch64-tdep.h"
>
> @@ -195,6 +196,29 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
> fprintf_filtered (file, _("AArch64 debugging is %s.\n"), value);
> }
>
> +/* Abstract instruction reader. */
> +
> +class abstract_instruction_reader
> +{
There is a new line before the class declaration here, but not on some
of the other declarations.
> +public:
> + /* Read in one instruction. */
> + virtual ULONGEST read (CORE_ADDR memaddr, int len,
> + enum bfd_endian byte_order) = 0;
> +};
> +
> +/* Instruction reader from real target. */
> +
> +class instruction_reader : public abstract_instruction_reader
> +{
> + public:
> + instruction_reader () = default;
> +
> + ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
> + {
> + return read_memory_unsigned_integer (memaddr, len, byte_order);
> + }
> +};
> +
> /* Analyze a prologue, looking for a recognizable stack frame
> and frame pointer. Scan until we encounter a store that could
> clobber the stack frame unexpectedly, or an unknown instruction. */
> @@ -202,7 +226,8 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
> static CORE_ADDR
> aarch64_analyze_prologue (struct gdbarch *gdbarch,
> CORE_ADDR start, CORE_ADDR limit,
> - struct aarch64_prologue_cache *cache)
> + struct aarch64_prologue_cache *cache,
> + abstract_instruction_reader& reader)
> {
> enum bfd_endian byte_order_for_code = gdbarch_byte_order_for_code (gdbarch);
> int i;
> @@ -221,7 +246,7 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> uint32_t insn;
> aarch64_inst inst;
>
> - insn = read_memory_unsigned_integer (start, 4, byte_order_for_code);
> + insn = reader.read (start, 4, byte_order_for_code);
>
> if (aarch64_decode_insn (insn, &inst, 1) != 0)
> break;
> @@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
> return start;
> }
>
> +static CORE_ADDR
> +aarch64_analyze_prologue (struct gdbarch *gdbarch,
> + CORE_ADDR start, CORE_ADDR limit,
> + struct aarch64_prologue_cache *cache)
> +{
> + instruction_reader reader { };
> +
> + return aarch64_analyze_prologue (gdbarch, start, limit, cache,
> + reader);
> +}
> +
> +#if GDB_SELF_TEST
> +
> +namespace selftests {
> +
> + /* Instruction reader from manually cooked instruction sequences. */
> + class instruction_reader_test : public abstract_instruction_reader
> + {
No newline between class declaration and comment.
> + public:
> + instruction_reader_test() = default ;
> + instruction_reader_test (std::initializer_list<uint32_t> init)
> + : insns{init} {}
> +
> + ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
> + {
> + SELF_CHECK (len == 4);
> + SELF_CHECK (memaddr % 4 == 0);
> + SELF_CHECK (memaddr / 4 < insns.size());
> +
> + return insns[memaddr / 4];
> + }
> +
> + private:
> + std::vector<uint32_t> insns;
> + };
Should the curly braces both be on their own lines as the rest of the
uses? I see mixed formatting in the uses of namespace we currently have.
We should pick one and go with it.
I particularly dislike the curly brace on the same line being used to
GNU's coding standards.
> +
> +static void
> +aarch64_analyze_prologue_test (void)
> +{
> + struct gdbarch_info info;
> +
> + gdbarch_info_init (&info);
> + info.bfd_arch_info = bfd_scan_arch ("aarch64");
> +
> + struct gdbarch *gdbarch = gdbarch_find_by_info (info);
> + SELF_CHECK (gdbarch != NULL);
> +
> + struct aarch64_prologue_cache cache;
> + cache.saved_regs = trad_frame_alloc_saved_regs (gdbarch);
> +
> + instruction_reader_test test {
> + 0xa9af7bfd, /* stp x29, x30, [sp,#-272]! */
> + 0x910003fd, /* mov x29, sp */
> + 0x97ffffe6, /* bl 0x400580 */
> + };
Same here.
> +
> + CORE_ADDR end = aarch64_analyze_prologue (gdbarch, 0, 128,
> + &cache, test);
> + SELF_CHECK (end == 4 * 2);
> +
> + SELF_CHECK (cache.framereg == AARCH64_FP_REGNUM);
> + SELF_CHECK (cache.framesize == 272);
> +
> + for (int i = 0; i < AARCH64_X_REGISTER_COUNT; i++)
> + {
> + if (i == AARCH64_FP_REGNUM)
> + SELF_CHECK (cache.saved_regs[i].addr == -272);
> + else if (i == AARCH64_LR_REGNUM)
> + SELF_CHECK (cache.saved_regs[i].addr == -264);
> + else
> + SELF_CHECK (cache.saved_regs[i].addr == -1);
> + }
> +
> + for (int i = 0; i < AARCH64_D_REGISTER_COUNT; i++)
> + {
> + int regnum = gdbarch_num_regs (gdbarch);
> +
> + SELF_CHECK (cache.saved_regs[i + regnum + AARCH64_D0_REGNUM].addr == -1);
> + }
> +}
> +}
> +#endif /* GDB_SELF_TEST */
> +
> /* Implement the "skip_prologue" gdbarch method. */
>
> static CORE_ADDR
> @@ -2864,6 +2972,10 @@ When on, AArch64 specific debugging is enabled."),
> NULL,
> show_aarch64_debug,
> &setdebuglist, &showdebuglist);
> +
> +#if GDB_SELF_TEST
> + register_self_test (selftests::aarch64_analyze_prologue_test);
> +#endif
> }
>
> /* AArch64 process record-replay related structures, defines etc. */
> diff --git a/gdb/trad-frame.c b/gdb/trad-frame.c
> index ebf19df..4430dd5 100644
> --- a/gdb/trad-frame.c
> +++ b/gdb/trad-frame.c
> @@ -43,16 +43,10 @@ trad_frame_cache_zalloc (struct frame_info *this_frame)
> return this_trad_cache;
> }
>
> -/* A traditional frame is unwound by analysing the function prologue
> - and using the information gathered to track registers. For
> - non-optimized frames, the technique is reliable (just need to check
> - for all potential instruction sequences). */
> -
> struct trad_frame_saved_reg *
> -trad_frame_alloc_saved_regs (struct frame_info *this_frame)
> +trad_frame_alloc_saved_regs (struct gdbarch *gdbarch)
> {
> int regnum;
> - struct gdbarch *gdbarch = get_frame_arch (this_frame);
> int numregs = gdbarch_num_regs (gdbarch) + gdbarch_num_pseudo_regs (gdbarch);
> struct trad_frame_saved_reg *this_saved_regs
> = FRAME_OBSTACK_CALLOC (numregs, struct trad_frame_saved_reg);
> @@ -65,6 +59,19 @@ trad_frame_alloc_saved_regs (struct frame_info *this_frame)
> return this_saved_regs;
> }
>
> +/* A traditional frame is unwound by analysing the function prologue
> + and using the information gathered to track registers. For
> + non-optimized frames, the technique is reliable (just need to check
> + for all potential instruction sequences). */
> +
> +struct trad_frame_saved_reg *
> +trad_frame_alloc_saved_regs (struct frame_info *this_frame)
> +{
> + struct gdbarch *gdbarch = get_frame_arch (this_frame);
> +
> + return trad_frame_alloc_saved_regs (gdbarch);
> +}
> +
> enum { TF_REG_VALUE = -1, TF_REG_UNKNOWN = -2 };
>
> int
> diff --git a/gdb/trad-frame.h b/gdb/trad-frame.h
> index b8aed16..d1c24b0 100644
> --- a/gdb/trad-frame.h
> +++ b/gdb/trad-frame.h
> @@ -104,6 +104,7 @@ int trad_frame_realreg_p (struct trad_frame_saved_reg this_saved_regs[],
>
> /* Return a freshly allocated (and initialized) trad_frame array. */
> struct trad_frame_saved_reg *trad_frame_alloc_saved_regs (struct frame_info *);
> +struct trad_frame_saved_reg *trad_frame_alloc_saved_regs (struct gdbarch *);
>
> /* Given the trad_frame info, return the location of the specified
> register. */
>
Otherwise looks good to me.
On 11/30/2016 07:30 PM, Luis Machado wrote:
> Should the curly braces both be on their own lines as the rest of the
> uses? I see mixed formatting in the uses of namespace we currently have.
> We should pick one and go with it.
>
> I particularly dislike the curly brace on the same line being used to
> GNU's coding standards.
Looking at the GCC codebase, the style with trailing { on the same
line (my preferred) is prevalent:
$ grep "^namespace " gcc/* | grep "{$" | wc -l
282
$ grep "^namespace " gcc/* | grep -v "{$" | wc -l
28
That style lends itself to opening nested namespaces more
like a FQN, like:
namespace foo { namespace bar {
With C++17, we'll be able to write:
namespace foo::bar {
Thanks,
Pedro Alves
@@ -44,6 +44,7 @@
#include "infcall.h"
#include "ax.h"
#include "ax-gdb.h"
+#include "selftest.h"
#include "aarch64-tdep.h"
@@ -195,6 +196,29 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
fprintf_filtered (file, _("AArch64 debugging is %s.\n"), value);
}
+/* Abstract instruction reader. */
+
+class abstract_instruction_reader
+{
+public:
+ /* Read in one instruction. */
+ virtual ULONGEST read (CORE_ADDR memaddr, int len,
+ enum bfd_endian byte_order) = 0;
+};
+
+/* Instruction reader from real target. */
+
+class instruction_reader : public abstract_instruction_reader
+{
+ public:
+ instruction_reader () = default;
+
+ ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
+ {
+ return read_memory_unsigned_integer (memaddr, len, byte_order);
+ }
+};
+
/* Analyze a prologue, looking for a recognizable stack frame
and frame pointer. Scan until we encounter a store that could
clobber the stack frame unexpectedly, or an unknown instruction. */
@@ -202,7 +226,8 @@ show_aarch64_debug (struct ui_file *file, int from_tty,
static CORE_ADDR
aarch64_analyze_prologue (struct gdbarch *gdbarch,
CORE_ADDR start, CORE_ADDR limit,
- struct aarch64_prologue_cache *cache)
+ struct aarch64_prologue_cache *cache,
+ abstract_instruction_reader& reader)
{
enum bfd_endian byte_order_for_code = gdbarch_byte_order_for_code (gdbarch);
int i;
@@ -221,7 +246,7 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
uint32_t insn;
aarch64_inst inst;
- insn = read_memory_unsigned_integer (start, 4, byte_order_for_code);
+ insn = reader.read (start, 4, byte_order_for_code);
if (aarch64_decode_insn (insn, &inst, 1) != 0)
break;
@@ -436,6 +461,89 @@ aarch64_analyze_prologue (struct gdbarch *gdbarch,
return start;
}
+static CORE_ADDR
+aarch64_analyze_prologue (struct gdbarch *gdbarch,
+ CORE_ADDR start, CORE_ADDR limit,
+ struct aarch64_prologue_cache *cache)
+{
+ instruction_reader reader { };
+
+ return aarch64_analyze_prologue (gdbarch, start, limit, cache,
+ reader);
+}
+
+#if GDB_SELF_TEST
+
+namespace selftests {
+
+ /* Instruction reader from manually cooked instruction sequences. */
+ class instruction_reader_test : public abstract_instruction_reader
+ {
+ public:
+ instruction_reader_test() = default ;
+ instruction_reader_test (std::initializer_list<uint32_t> init)
+ : insns{init} {}
+
+ ULONGEST read (CORE_ADDR memaddr, int len, enum bfd_endian byte_order)
+ {
+ SELF_CHECK (len == 4);
+ SELF_CHECK (memaddr % 4 == 0);
+ SELF_CHECK (memaddr / 4 < insns.size());
+
+ return insns[memaddr / 4];
+ }
+
+ private:
+ std::vector<uint32_t> insns;
+ };
+
+static void
+aarch64_analyze_prologue_test (void)
+{
+ struct gdbarch_info info;
+
+ gdbarch_info_init (&info);
+ info.bfd_arch_info = bfd_scan_arch ("aarch64");
+
+ struct gdbarch *gdbarch = gdbarch_find_by_info (info);
+ SELF_CHECK (gdbarch != NULL);
+
+ struct aarch64_prologue_cache cache;
+ cache.saved_regs = trad_frame_alloc_saved_regs (gdbarch);
+
+ instruction_reader_test test {
+ 0xa9af7bfd, /* stp x29, x30, [sp,#-272]! */
+ 0x910003fd, /* mov x29, sp */
+ 0x97ffffe6, /* bl 0x400580 */
+ };
+
+ CORE_ADDR end = aarch64_analyze_prologue (gdbarch, 0, 128,
+ &cache, test);
+ SELF_CHECK (end == 4 * 2);
+
+ SELF_CHECK (cache.framereg == AARCH64_FP_REGNUM);
+ SELF_CHECK (cache.framesize == 272);
+
+ for (int i = 0; i < AARCH64_X_REGISTER_COUNT; i++)
+ {
+ if (i == AARCH64_FP_REGNUM)
+ SELF_CHECK (cache.saved_regs[i].addr == -272);
+ else if (i == AARCH64_LR_REGNUM)
+ SELF_CHECK (cache.saved_regs[i].addr == -264);
+ else
+ SELF_CHECK (cache.saved_regs[i].addr == -1);
+ }
+
+ for (int i = 0; i < AARCH64_D_REGISTER_COUNT; i++)
+ {
+ int regnum = gdbarch_num_regs (gdbarch);
+
+ SELF_CHECK (cache.saved_regs[i + regnum + AARCH64_D0_REGNUM].addr == -1);
+ }
+}
+}
+#endif /* GDB_SELF_TEST */
+
/* Implement the "skip_prologue" gdbarch method. */
static CORE_ADDR
@@ -2864,6 +2972,10 @@ When on, AArch64 specific debugging is enabled."),
NULL,
show_aarch64_debug,
&setdebuglist, &showdebuglist);
+
+#if GDB_SELF_TEST
+ register_self_test (selftests::aarch64_analyze_prologue_test);
+#endif
}
/* AArch64 process record-replay related structures, defines etc. */
@@ -43,16 +43,10 @@ trad_frame_cache_zalloc (struct frame_info *this_frame)
return this_trad_cache;
}
-/* A traditional frame is unwound by analysing the function prologue
- and using the information gathered to track registers. For
- non-optimized frames, the technique is reliable (just need to check
- for all potential instruction sequences). */
-
struct trad_frame_saved_reg *
-trad_frame_alloc_saved_regs (struct frame_info *this_frame)
+trad_frame_alloc_saved_regs (struct gdbarch *gdbarch)
{
int regnum;
- struct gdbarch *gdbarch = get_frame_arch (this_frame);
int numregs = gdbarch_num_regs (gdbarch) + gdbarch_num_pseudo_regs (gdbarch);
struct trad_frame_saved_reg *this_saved_regs
= FRAME_OBSTACK_CALLOC (numregs, struct trad_frame_saved_reg);
@@ -65,6 +59,19 @@ trad_frame_alloc_saved_regs (struct frame_info *this_frame)
return this_saved_regs;
}
+/* A traditional frame is unwound by analysing the function prologue
+ and using the information gathered to track registers. For
+ non-optimized frames, the technique is reliable (just need to check
+ for all potential instruction sequences). */
+
+struct trad_frame_saved_reg *
+trad_frame_alloc_saved_regs (struct frame_info *this_frame)
+{
+ struct gdbarch *gdbarch = get_frame_arch (this_frame);
+
+ return trad_frame_alloc_saved_regs (gdbarch);
+}
+
enum { TF_REG_VALUE = -1, TF_REG_UNKNOWN = -2 };
int
@@ -104,6 +104,7 @@ int trad_frame_realreg_p (struct trad_frame_saved_reg this_saved_regs[],
/* Return a freshly allocated (and initialized) trad_frame array. */
struct trad_frame_saved_reg *trad_frame_alloc_saved_regs (struct frame_info *);
+struct trad_frame_saved_reg *trad_frame_alloc_saved_regs (struct gdbarch *);
/* Given the trad_frame info, return the location of the specified
register. */