mailing list of musl libc
 help / color / mirror / code / Atom feed
From: Szabolcs Nagy <nsz@port70.net>
To: musl@lists.openwall.com
Subject: Re: [PATCH] First prototype of script which adds CFI directives to x86 asm
Date: Wed, 13 May 2015 13:39:18 +0200	[thread overview]
Message-ID: <20150513113918.GE31118@port70.net> (raw)
In-Reply-To: <1431466124-2848-1-git-send-email-alexinbeijing@gmail.com>

* Alex Dowad <alexinbeijing@gmail.com> [2015-05-12 23:28:44 +0200]:
> diff --git a/Makefile b/Makefile
> index 6559295..f7335aa 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -118,7 +118,11 @@ $(foreach s,$(wildcard src/*/$(ARCH)*/*.s),$(eval $(call mkasmdep,$(s))))
>  	$(CC) $(CFLAGS_ALL_STATIC) -c -o $@ $(dir $<)$(shell cat $<)
>  
>  %.o: $(ARCH)/%.s
> +ifeq ($(ARCH),i386)
> +	awk -f tools/add-cfi.awk.i386 $< | $(CC) $(CFLAGS_ALL_STATIC) -x assembler -c -o $@ -
> +else
>  	$(CC) $(CFLAGS_ALL_STATIC) -c -o $@ $<
> +endif
>  

it might make sense to have a make rule that produces the cfi asm for inspection

%.cfi.s: $(ARCH)/%.s
	$(ADD_CFI) $< > $@
%.o: %.cfi.s
	$(CC) $(CFLAGS_ALL_STATIC) -c -o $@ $<

(unless ppl feel that adds too much clutter.. the clean rule should be
adjusted too, but make sometimes removes the intermediates based on some
mysterious logic)

i wonder if a configure check for .cfi support should be added: in theory
an assembler may not support it (tcc?)

> +++ b/tools/add-cfi.awk.i386
> @@ -0,0 +1,152 @@
> +# Insert GAS CFI directives ("control frame information") into x86-32 asm input
> +#
> +# CFI directives tell the assembler how to generate "stack frame" debug info
> +# This information can tell a debugger (like gdb) how to find the current stack
> +#   frame at any point in the program code, and how to find the values which
> +#   various registers had at higher points in the call stack
> +# With this information, the debugger can show a backtrace, and you can move up
> +#   and down the call stack and examine the values of local variables
> +
> +BEGIN {
> +  # don"t put CFI data in the .eh_frame ELF section (which we don't keep)
> +  print ".cfi_sections .debug_frame"
> +
> +  # only emit CFI directives inside a function callable from C
> +  # (blindly emitting a '.cfi_startproc' at the beginning of each file and
> +  #   '.cfi_endproc' at the end doesn't work)
> +  in_function = 0
> +}
> +
> +function hex2int(str) {
> +  str = tolower(str)
> +
> +  for (i = 1; i <= 16; i++) {
> +    char = substr("0123456789abcdef", i, 1)
> +    lookup[char] = i-1
> +  }
> +

move this loop to BEGIN so it only runs at startup

(in a more complex script you should mark local variables otherwise
everything is in the global namespace and function calls clobber each
other's temporaries.. here it is ok as there arent many nested calls,
but i tend to add variables like "i" to the argument list)

> +  result = 0
> +  for (i = 1; i <= length(str); i++) {
> +    result = result * 16
> +    char   = substr(str, i, 1)
> +    result = result + lookup[char]
> +  }
> +  return result
> +}
> +
> +function get_const() {
> +  # only use if you already know there is 1 and only 1 constant
> +  match($0, /\$[0-9a-fA-F]+/)
> +  return hex2int(substr($0, RSTART+1, RLENGTH-1))
> +}

i think hex conversion for $123 is wrong in i386 asm

you should only hex convert $0x123

> +function get_reg() {
> +  # only use if you already know there is 1 and only 1 register
> +  match($0, /%e(ax|bx|cx|dx|si|di|bp)/)
> +  return substr($0, RSTART+1, RLENGTH-1)
> +}
> +function get_reg1() {
> +  # for instructions with 2 operands, get 1st operand (assuming it is register)
> +  match($0, /%e(ax|bx|cx|dx|si|di|bp),/)
> +  return substr($0, RSTART+1, RLENGTH-2)
> +}
> +function get_reg2() {
> +  # for instructions with 2 operands, get 2nd operand (assuming it is register)
> +  match($0, /,%e(ax|bx|cx|dx|si|di|bp)/)
> +  return substr($0, RSTART+2, RLENGTH-2)

allow whitespace between ',' and the regs

(or even better: have a global rule to canonicalize the code
removing whitespace, comments etc, but be careful about constants
eg. avoid changing .ascii "foo")

http://sourceware.org/binutils/docs/as/Constants.html

> +}
> +
> +function adjust_sp_offset(delta) {
> +  if (in_function) {
> +    printf ".cfi_adjust_cfa_offset %d\n", delta
> +  }
> +}
> +
> +{ print }
> +
> +/\.type.*,@function/ {
> +  if (in_function) {
> +    print ".cfi_endproc"
> +  }
> +

(currently this will match inside comments)

i noticed that crt/ asm does not have .type directives
i wonder if that's intentional

(missing .cfi_startproc/endproc might be problematic i think
because .cfi directives can be rejected outside of startproc/endproc)

note that there are functions with aliases where you have

.global foo
.global bar
.type foo,@function
.type bar,@function
foo:
bar:
	... code

so there will be an empty startproc/endproc there..
not sure if that causes any problems

> +  print ".cfi_startproc"
> +  in_function = 1
> +
> +  for (register in saved)
> +    delete saved[register]
> +  for (register in dirty)
> +    delete dirty[register]
> +}
> +
> +# KEEPING UP WITH THE STACK POINTER
> +# We do NOT attempt to understand foolish and ridiculous tricks like stashing
> +#   the stack pointer and then using %esp as a scratch register, or bitshifting
> +#   it or taking its square root or anything stupid like that.
> +# %esp should only be adjusted by pushing/popping or adding/subtracting constants
> +#
> +/pushl?/ { adjust_sp_offset(4) }
> +/popl?/  { adjust_sp_offset(-4) }

i think it's possible to push 2 bytes (push %ax)

> +# TODO: can add/sub instructions also specify offset in decimal?

yes :)

> +# TODO: can offset be negative?

yes $-123 is a valid constant

(it should be the same as $0xffffff85 in 32bit contexts)

> +/addl?\s+\$[0-9a-fA-F]+,%esp/ { adjust_sp_offset(-get_const()) }
> +/subl?\s+\$[0-9a-fA-F]+,%esp/ { adjust_sp_offset(get_const()) }
> +
> +# TRACKING REGISTER VALUES FROM THE PREVIOUS STACK FRAME
> +#
> +/pushl?\s+%e(ax|bx|cx|dx|si|di|bp)/ { # don't match "push (%reg)"
> +  # if a register is being pushed, and its value has not changed since the
> +  #   beginning of this function, the pushed value can be used when printing
> +  #   local variables at the next level up the stack
> +  # emit '.cfi_rel_offset' for that
> +
> +  if (in_function) {
> +    register = get_reg()
> +    if (!saved[register] && !dirty[register]) {
> +      printf ".cfi_rel_offset %s,0\n", register
> +      saved[register] = 1
> +    }
> +  }
> +}
> +
> +# TODO: this should also understand hex offsets prefixed with 0x or -0x
> +/movl?\s+%e(ax|bx|cx|dx|si|di|bp),-?[0-9]*\(%esp\)/ {

yes, i think it can be hex

> +  if (in_function) {
> +    register = get_reg()
> +    if (match($0, /-?[0-9]+\(%esp\)/)) {
> +      offset = substr($0, RSTART, RLENGTH-6) # decimal, not hex!
> +    } else {
> +      offset = 0
> +    }
> +    if (!saved[register] && !dirty[register]) {
> +      printf ".cfi_rel_offset %s,%d\n", register, offset
> +      saved[register] = 1
> +    }
> +  }
> +}
> +
> +# IF REGISTER VALUES ARE UNCEREMONIOUSLY TRASHED
> +# ...then we want to know about it.
> +#
> +function trashed(register) {
> +  if (in_function && !saved[register] && !dirty[register]) {
> +    printf ".cfi_undefined %s\n", register
> +  }
> +  dirty[register] = 1
> +}
> +# this does NOT exhaustively check for all possible instructions which could
> +# overwrite a register value inherited from the caller (just the common ones)
> +# TODO: detect when ax/ah/al/etc. are trashed -- means eax is no longer usable either
> +/mov.*,%e(ax|bx|cx|dx|si|di|bp)/  { trashed(get_reg2()) }
> +/(add|addl|sub|subl|and|or|xor|lea|sal|sar|shl|shr)\s+%e(ax|bx|cx|dx|si|di|bp),/ {
> +  trashed(get_reg1())
> +}
> +/i?mul\s+[^,]*$/                    { trashed("eax"); trashed("edx") }
> +/i?mul\s+%e(ax|bx|cx|dx|si|di|bp),/ { trashed(get_reg1()) }
> +/^(\w+:)?\s*i?div/                  { trashed("eax"); trashed("edx") }
> +/(dec|inc|not|neg|pop)\s+%e(ax|bx|cx|dx|si|di|bp)/  { trashed(get_reg()) }
> +/^(\w+:)\s*cpuid/ { trashed("eax"); trashed("ebx"); trashed("ecx"); trashed("edx") }
> +
> +END {
> +  if (in_function) {
> +    print ".cfi_endproc"
> +  }
> +}
> \ No newline at end of file
> -- 
> 2.0.0.GIT


      parent reply	other threads:[~2015-05-13 11:39 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-12 21:28 Alex Dowad
2015-05-12 22:52 ` Rich Felker
2015-05-13 11:39 ` Szabolcs Nagy [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20150513113918.GE31118@port70.net \
    --to=nsz@port70.net \
    --cc=musl@lists.openwall.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.vuxu.org/mirror/musl/

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).