| Message ID | 5798785F.8020001@redhat.com |
|---|---|
| State | New |
| Headers | show |
Hi Aldy, > Just in case this got lost in noise, since I know there was a lot of back > and forth between Martin Sebor and I. > > This is the last iteration. > > Tested on x86-64 Linux. > > OK for trunk? > > gcc/ > > * Makefile.in (OBJS): Add gimple-ssa-warn-walloca.o. > * passes.def: Add two instances of pass_walloca. > * tree-pass.h (make_pass_walloca): New. > * gimple-ssa-warn-walloca.c: New file. just a nit: the files is called gimple-ssa-warn-alloca.[co]. Rainer
On 07/27/2016 03:01 AM, Aldy Hernandez wrote: > Just in case this got lost in noise, since I know there was a lot of > back and forth between Martin Sebor and I. > > This is the last iteration. > > Tested on x86-64 Linux. > > OK for trunk? > > curr > > > gcc/ > > * Makefile.in (OBJS): Add gimple-ssa-warn-walloca.o. > * passes.def: Add two instances of pass_walloca. > * tree-pass.h (make_pass_walloca): New. > * gimple-ssa-warn-walloca.c: New file. > * doc/invoke.texi: Document -Walloca, -Walloca-larger-than=, and > -Wvla-larger-than= options. > > gcc/c-family/ > > * c.opt (Walloca): New. > (Walloca-larger-than=): New. > (Wvla-larger-than=): New. As someone already noted, it's gimple-ssa-warn-alloca, not gimple-ssa-warn-walloca for the ChangeLog entry. On the nittish side, you're mixing C and C++ comment styles. Choosing one and sticking with it seems better :-) > > +@item -Walloca > +@opindex Wno-alloca > +@opindex Walloca > +This option warns on all uses of @code{alloca} in the source. > + > +@item -Walloca-larger-than=@var{n} > +This option warns on calls to @code{alloca} that are not bounded by a > +controlling predicate limiting its size to @var{n} bytes, or calls to > +@code{alloca} where the bound is unknown. So for each of these little examples, I'd stuff the code into a trivial function definition and make "n" a parameter. That way it's obvious the value of "n" comes from a context where we don't initially know its range, but we may be able to narrow the range due to statements in the function. ; > + > +class pass_walloca : public gimple_opt_pass > +{ > +public: > + pass_walloca (gcc::context *ctxt) > + : gimple_opt_pass(pass_data_walloca, ctxt), first_time_p (false) > + {} > + opt_pass *clone () { return new pass_walloca (m_ctxt); } > + void set_pass_param (unsigned int n, bool param) > + { > + gcc_assert (n == 0); > + first_time_p = param; > + } ISTM that you're using "first_time_p" here, but in passes.def you refer to this parameter as "strict_mode_p" in comments. ie: + NEXT_PASS (pass_walloca, /*strict_mode_p=*/false); I'd just drop the /*strict_mode_p*/ comment in both places it appears in your patch's change to passes.def. I think we've generally frowned on those embedded comments, even though some have snuck in. > + > +// We have a few heuristics up our sleeve to determine if a call to > +// alloca() is within bounds. Try them out and return the type of > +// alloca call this is based on its argument. > +// > +// Given a known argument (ARG) to alloca() and an EDGE (E) > +// calculating said argument, verify that the last statement in the BB > +// in E->SRC is a gate comparing ARG to an acceptable bound for > +// alloca(). See examples below. > +// > +// MAX_SIZE is WARN_ALLOCA= adjusted for VLAs. It is the maximum size > +// in bytes we allow for arg. > +// > +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is > +// set to the bound used to determine this. ASSUMED_LIMIT is only set > +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. > +// > +// Returns the alloca type. > + > +static enum alloca_type > +alloca_call_type_by_arg (tree arg, edge e, unsigned max_size, > + wide_int *assumed_limit) So I wonder if you ought to have a structure here for the return value which contains the alloca type and assumed limit. I know in the past we avoided aggregate returns, but these days that doesn't seem necessary. Seems cleaner than having a return value and output parameters. > +{ > + // All the tests bellow depend on the jump being on the TRUE path. > + if (!(e->flags & EDGE_TRUE_VALUE)) > + return ALLOCA_UNBOUNDED; Seems like a fairly arbitrary and undesirable limitation. Couldn't the developer just have easily written if (arg > N> x = malloc (...) else x = alloca (...) It also seems like you'd want to handle the set of LT/LE/GT/GE rather than just LE. Or is it the case that we always canonicalize LT into LE by adjusting the constant (I vaguely remember running into that in RTL, so it's entirely possible and there'd likely be a canonicalization of GT/GE as well). It also seems that once Andrew's infrastructure is in place this becomes dead code as we can just ask for the range at a point in the program, including for each incoming edge. You might want a comment to that effect. > + > + /* Check for: > + if (arg .cond. LIMIT) -or- if (LIMIT .cond. arg) > + alloca(arg); > + > + Where LIMIT has a bound of unknown range. */ > + tree limit = NULL; > + if (gimple_cond_lhs (last) == arg) > + limit = gimple_cond_rhs (last); > + else if (gimple_cond_rhs (last) == arg) > + limit = gimple_cond_lhs (last); > + if (limit && TREE_CODE (limit) == SSA_NAME) > + { > + wide_int min, max; > + value_range_type range_type = get_range_info (limit, &min, &max); > + if (range_type == VR_UNDEFINED || range_type == VR_VARYING) > + return ALLOCA_BOUND_UNKNOWN; > + // FIXME: We could try harder here and handle a possible range > + // or anti-range. Hopefully the upcoming changes to range info > + // will give us finer grained info, and we can avoid somersaults > + // here. Ah, can't you set *assumed_limit here? It's just a matter of walking through the cases and making the most conservative assumption. So assume the condition is LT, both objects are unsigned types and LIMIT has a range like [5..25]. Then the resulting *assumed_limit must be 24. ISTM it might also be worth checking VRP here -- I'd expect it to be able to make this kind of determination. that would be independent of this work (in the sense that if VRP isn't creating ranges for this, it should be fixed independently). > + } > + > + return ALLOCA_UNBOUNDED; > +} > + > +// Return TRUE if SSA's definition is a cast from a signed type. > +// If so, set *INVALID_CASTED_TYPE to the signed type. > + > +static bool > +cast_from_signed_p (tree ssa, tree *invalid_casted_type) > +{ > + gimple *def = SSA_NAME_DEF_STMT (ssa); > + if (def > + && !gimple_nop_p (def) > + && gimple_assign_cast_p (def) > + && !TYPE_UNSIGNED (TREE_TYPE (gimple_assign_rhs1 (def)))) > + { > + *invalid_casted_type = TREE_TYPE (gimple_assign_rhs1 (def)); > + return true; > + } > + return false; Note that we may have a cast from a signed type, but if the RHS of that cast has a positive range, then the cast isn't going to case the wrap-around effect that is so problematical. That might help cut down false positives. > +} > + > +// Return TURE if X has a maximum range of MAX, basically covering the > +// entire domain, in which case it's no range at all. s/TURE/TRUE/ > + > +static bool > +is_max (tree x, wide_int max) > +{ > + return wi::max_value (TREE_TYPE (x)) == max; > +} I'm a bit surprised we don't have this kind of utility function elsewhere. I wonder if it'd be better to conceptualize this as a range query since it looks like you're always using get_range_info to get an object's range, then comparing what that returns to the maximal value of hte object's type. Maybe that's too much bikeshedding... > + > +// Analyze the alloca call in STMT and return an `enum alloca_type' > +// explaining what type of alloca it is. IS_VLA is set if the alloca > +// call is really a BUILT_IN_ALLOCA_WITH_ALIGN, signifying a VLA. > +// > +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is > +// set to the bound used to determine this. ASSUMED_LIMIT is only set > +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. > +// > +// If the alloca call may be too large because of a cast from a signed > +// type to an unsigned type, set *INVALID_CASTED_TYPE to the > +// problematic signed type. > + > +static enum alloca_type > +alloca_call_type (gimple *stmt, bool is_vla, wide_int *assumed_limit, > + tree *invalid_casted_type) Again, consider an aggregate return. > +{ > + gcc_assert (gimple_alloca_call_p (stmt)); > + tree len = gimple_call_arg (stmt, 0); > + enum alloca_type w = ALLOCA_UNBOUNDED; > + wide_int min, max; > + > + gcc_assert (!is_vla || warn_vla_limit > 0); > + gcc_assert (is_vla || warn_alloca_limit > 0); > + > + // Adjust warn_alloca_max_size for VLAs, by taking the underlying > + // type into account. > + unsigned HOST_WIDE_INT max_size; > + if (is_vla) > + max_size = (unsigned HOST_WIDE_INT) warn_vla_limit; > + else > + max_size = (unsigned HOST_WIDE_INT) warn_alloca_limit; > + > + // Check for the obviously bounded case. > + if (TREE_CODE (len) == INTEGER_CST) > + { > + if (tree_to_uhwi (len) > max_size) > + { > + *assumed_limit = len; > + return ALLOCA_BOUND_DEFINITELY_LARGE; > + } > + if (integer_zerop (len)) > + return ALLOCA_ARG_IS_ZERO; > + w = ALLOCA_OK; > + } > + else if (TREE_CODE (len) != SSA_NAME) > + return ALLOCA_UNBOUNDED; Hmm, other than INTEGER_CST and SSA_NAME, is there any other nodes we can get here? Perhaps we get DECLs and such, particularly when not optimizing?!? > + // Check the range info if available. > + else > + { > + if (value_range_type range_type = get_range_info (len, &min, &max)) > + { > + if (range_type == VR_RANGE) > + { > + if (wi::leu_p (max, max_size)) > + w = ALLOCA_OK; > + else if (is_max (len, max)) > + { > + // A cast may have created a range we don't care > + // about. For instance, a cast from 16-bit to > + // 32-bit creates a range of 0..65535, even if there > + // is not really a determinable range in the > + // underlying code. In this case, look through the > + // cast at the original argument, and fall through > + // to look at other alternatives. > + gimple *def = SSA_NAME_DEF_STMT (len); > + if (gimple_assign_cast_p (def)) > + len = gimple_assign_rhs1 (def); > + } > + else > + { > + /* If `len' is merely a cast that is being > + calculated right before the call to alloca, look > + at the range for the original value. Yea, this is similar to my comment earlier that the RHS of the cast may have a known range (non-negative) that allows us to not worry about the cast creating a huge integer. I think you can add handling that case here without too much trouble. Though you might consider pulling all the casting bits into a separate function. > + > + This avoids the cast creating a range where the > + original expression did not have a range: > + > + # RANGE [0, 18446744073709551615] NONZERO 4294967295 > + _2 = (long unsigned int) n_7(D); > + p_9 = __builtin_alloca (_2); Note this example would make more sense if the type of n_7 was explicit. > + else if (range_type == VR_ANTI_RANGE) > + { > + // There may be some wrapping around going on. Catch it > + // with this heuristic. Hopefully, this VR_ANTI_RANGE > + // nonsense will go away, and we won't have to catch the > + // sign conversion problems with this crap. > + if (cast_from_signed_p (len, invalid_casted_type)) > + return ALLOCA_CAST_FROM_SIGNED; Another place where casting from an object with a non-negative range ought to be filtered out as not problematical. > + > + // If we couldn't find anything, try a few heuristics for things we > + // can easily determine. Check these misc cases but only accept > + // them if all predecessors have a known bound. > + basic_block bb = gimple_bb (stmt); > + if (w == ALLOCA_UNBOUNDED) > + { > + w = ALLOCA_OK; > + for (unsigned ix = 0; ix < EDGE_COUNT (bb->preds); ix++) > + { > + enum alloca_type w > + = alloca_call_type_by_arg (len, EDGE_PRED (bb, ix), max_size, > + assumed_limit); > + if (w != ALLOCA_OK) > + return w; > + } So this is an interesting tidbit that answers my questions about how we use alloca_call_type_by_arg. Essentially this is meant to catch the merge point for flow control and give a conservative warning. That's fine and good. But ISTM it's really a bit of a hack. What if we have something like this: X Y Z \ | / \|/ A / \ / \ B C / \ / \ D E Where the alloca call is in E and the incoming edges to A actually have useful information about the argument to the alloca call. ISTM you need to be doing something with the dominator tree here to find the merge point(s) where we might know something useful. And it's this kind of test that makes me wonder about re-purposing some of the path analysis code from tree-ssa-uninit.c. It may be the case that the path Z->A->C->E is unfeasible, but left in the CFG because duplication to expose the unfeasible path was unprofitable. If it turns out that the only argument that causes problems comes from the edge Z->A, then we wouldn't want to warn in this case. I don't see Andrew's work necessarily being able to solve this problem. > + } > + > + return w; > +} > + > +// Return TRUE if the alloca call in STMT is in a loop, otherwise > +// return FALSE. As an exception, ignore alloca calls for VLAs that > +// occur in a loop since those will be cleaned up when they go out of > +// scope. > + > +static bool > +in_loop_p (bool is_vla, gimple *stmt) > +{ > + basic_block bb = gimple_bb (stmt); > + if (bb->loop_father > + // ?? Functions with no loops get a loop_father? I > + // don't get it. The following conditional seems to do > + // the trick to exclude such nonsense. > + && bb->loop_father->header != ENTRY_BLOCK_PTR_FOR_FN (cfun)) I believe there is a "loop" that encompasses the whole function. I'll probably have more comments after the next iteration. I'm also real curious what you've found poking around at GDB with this warning :-) Jeff
On 08/19/2016 05:35 AM, Aldy Hernandez wrote: > On 08/04/2016 12:37 PM, Jeff Law wrote: >> On 07/27/2016 03:01 AM, Aldy Hernandez wrote: >>> Just in case this got lost in noise, since I know there was a lot of >>> back and forth between Martin Sebor and I. >>> >>> This is the last iteration. >>> >>> Tested on x86-64 Linux. >>> >>> OK for trunk? >>> >>> curr >>> >>> >>> gcc/ >>> >>> * Makefile.in (OBJS): Add gimple-ssa-warn-walloca.o. >>> * passes.def: Add two instances of pass_walloca. >>> * tree-pass.h (make_pass_walloca): New. >>> * gimple-ssa-warn-walloca.c: New file. >>> * doc/invoke.texi: Document -Walloca, -Walloca-larger-than=, and >>> -Wvla-larger-than= options. >>> >>> gcc/c-family/ >>> >>> * c.opt (Walloca): New. >>> (Walloca-larger-than=): New. >>> (Wvla-larger-than=): New. >> As someone already noted, it's gimple-ssa-warn-alloca, not >> gimple-ssa-warn-walloca for the ChangeLog entry. > > Fixed. > >> >> On the nittish side, you're mixing C and C++ comment styles. Choosing >> one and sticking with it seems better :-) > > Fixed. Settled for C++ comments, except the copyright headers and the > testcases. > >> >> >>> >>> +@item -Walloca >>> +@opindex Wno-alloca >>> +@opindex Walloca >>> +This option warns on all uses of @code{alloca} in the source. >>> + >>> +@item -Walloca-larger-than=@var{n} >>> +This option warns on calls to @code{alloca} that are not bounded by a >>> +controlling predicate limiting its size to @var{n} bytes, or calls to >>> +@code{alloca} where the bound is unknown. >> So for each of these little examples, I'd stuff the code into a trivial >> function definition and make "n" a parameter. That way it's obvious the >> value of "n" comes from a context where we don't initially know its >> range, but we may be able to narrow the range due to statements in the >> function. > > Done. > >> >> ; >>> + >>> +class pass_walloca : public gimple_opt_pass >>> +{ >>> +public: >>> + pass_walloca (gcc::context *ctxt) >>> + : gimple_opt_pass(pass_data_walloca, ctxt), first_time_p (false) >>> + {} >>> + opt_pass *clone () { return new pass_walloca (m_ctxt); } >>> + void set_pass_param (unsigned int n, bool param) >>> + { >>> + gcc_assert (n == 0); >>> + first_time_p = param; >>> + } >> ISTM that you're using "first_time_p" here, but in passes.def you refer >> to this parameter as "strict_mode_p" in comments. >> >> ie: >> >> + NEXT_PASS (pass_walloca, /*strict_mode_p=*/false); >> >> I'd just drop the /*strict_mode_p*/ comment in both places it appears in >> your patch's change to passes.def. I think we've generally frowned on >> those embedded comments, even though some have snuck in. > > I've seen a lot of embedded comments throughout GCC, especially in > optional type arguments. ISTM it makes things clearer for these > parameters. But hey, I don't care that much. Fixed. > >> >>> + >>> +// We have a few heuristics up our sleeve to determine if a call to >>> +// alloca() is within bounds. Try them out and return the type of >>> +// alloca call this is based on its argument. >>> +// >>> +// Given a known argument (ARG) to alloca() and an EDGE (E) >>> +// calculating said argument, verify that the last statement in the BB >>> +// in E->SRC is a gate comparing ARG to an acceptable bound for >>> +// alloca(). See examples below. >>> +// >>> +// MAX_SIZE is WARN_ALLOCA= adjusted for VLAs. It is the maximum size >>> +// in bytes we allow for arg. >>> +// >>> +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is >>> +// set to the bound used to determine this. ASSUMED_LIMIT is only set >>> +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. >>> +// >>> +// Returns the alloca type. >>> + >>> +static enum alloca_type >>> +alloca_call_type_by_arg (tree arg, edge e, unsigned max_size, >>> + wide_int *assumed_limit) >> So I wonder if you ought to have a structure here for the return value >> which contains the alloca type and assumed limit. I know in the past we >> avoided aggregate returns, but these days that doesn't seem necessary. >> Seems cleaner than having a return value and output parameters. > > Done, C++ style with a simple constructor :). > >> >>> +{ >>> + // All the tests bellow depend on the jump being on the TRUE path. >>> + if (!(e->flags & EDGE_TRUE_VALUE)) >>> + return ALLOCA_UNBOUNDED; >> Seems like a fairly arbitrary and undesirable limitation. Couldn't the >> developer just have easily written >> >> if (arg > N> >> x = malloc (...) >> else >> x = alloca (...) >> >> It also seems like you'd want to handle the set of LT/LE/GT/GE rather >> than just LE. Or is it the case that we always canonicalize LT into LE >> by adjusting the constant (I vaguely remember running into that in RTL, >> so it's entirely possible and there'd likely be a canonicalization of >> GT/GE as well). > > Most of it gets canonicalized, but your testcase is definitely possible, > so I fixed this. > >> >> It also seems that once Andrew's infrastructure is in place this becomes >> dead code as we can just ask for the range at a point in the program, >> including for each incoming edge. You might want a comment to that >> effect. > > Done. > >> >> >> >> >>> + >>> + /* Check for: >>> + if (arg .cond. LIMIT) -or- if (LIMIT .cond. arg) >>> + alloca(arg); >>> + >>> + Where LIMIT has a bound of unknown range. */ >>> + tree limit = NULL; >>> + if (gimple_cond_lhs (last) == arg) >>> + limit = gimple_cond_rhs (last); >>> + else if (gimple_cond_rhs (last) == arg) >>> + limit = gimple_cond_lhs (last); >>> + if (limit && TREE_CODE (limit) == SSA_NAME) >>> + { >>> + wide_int min, max; >>> + value_range_type range_type = get_range_info (limit, &min, &max); >>> + if (range_type == VR_UNDEFINED || range_type == VR_VARYING) >>> + return ALLOCA_BOUND_UNKNOWN; >>> + // FIXME: We could try harder here and handle a possible range >>> + // or anti-range. Hopefully the upcoming changes to range info >>> + // will give us finer grained info, and we can avoid somersaults >>> + // here. >> Ah, can't you set *assumed_limit here? It's just a matter of walking >> through the cases and making the most conservative assumption. So >> assume the condition is LT, both objects are unsigned types and LIMIT >> has a range like [5..25]. Then the resulting *assumed_limit must be 24. > > Well, it turns out that we don't ever hit the FIXME path. We either > handle limits we know with the previous code block (which gets > normalized to [arg .cond. LIMIT] always, or we handle the unknown path > with this block with [LIMIT .cond. arg]. > > I've simplified the code a bit, and updated the comments. I also added > a gcc_unreachable() just in case we ever hit this path. Though I've > verified at least building glibc and binutils that we never do. > >> >> ISTM it might also be worth checking VRP here -- I'd expect it to be >> able to make this kind of determination. that would be independent of >> this work (in the sense that if VRP isn't creating ranges for this, it >> should be fixed independently). >> >> >>> + } >>> + >>> + return ALLOCA_UNBOUNDED; >>> +} >>> + >>> +// Return TRUE if SSA's definition is a cast from a signed type. >>> +// If so, set *INVALID_CASTED_TYPE to the signed type. >>> + >>> +static bool >>> +cast_from_signed_p (tree ssa, tree *invalid_casted_type) >>> +{ >>> + gimple *def = SSA_NAME_DEF_STMT (ssa); >>> + if (def >>> + && !gimple_nop_p (def) >>> + && gimple_assign_cast_p (def) >>> + && !TYPE_UNSIGNED (TREE_TYPE (gimple_assign_rhs1 (def)))) >>> + { >>> + *invalid_casted_type = TREE_TYPE (gimple_assign_rhs1 (def)); >>> + return true; >>> + } >>> + return false; >> Note that we may have a cast from a signed type, but if the RHS of that >> cast has a positive range, then the cast isn't going to case the >> wrap-around effect that is so problematical. That might help cut down >> false positives. > > Actually when the cast is from a known positive range we don't get a > VR_ANTI_RANGE, we get a proper VR_RANGE. > > I've cleaned this code up a bit and merged some common conditionals. In > the process, taking a subset of your advice, and cleaning up some things > I've managed to handle 2 cases where I was previously XFAILing. > So...less false positives. More coverage. Woo hoo! > >> >>> +} >>> + >>> +// Return TURE if X has a maximum range of MAX, basically covering the >>> +// entire domain, in which case it's no range at all. >> s/TURE/TRUE/ > > Fixed. > >> >> >>> + >>> +static bool >>> +is_max (tree x, wide_int max) >>> +{ >>> + return wi::max_value (TREE_TYPE (x)) == max; >>> +} >> I'm a bit surprised we don't have this kind of utility function >> elsewhere. I wonder if it'd be better to conceptualize this as a range >> query since it looks like you're always using get_range_info to get an >> object's range, then comparing what that returns to the maximal value of >> hte object's type. Maybe that's too much bikeshedding... > > *shrug*. If you feel strongly, I can look into this, but I'm inherently > lazy :). > >> >> >> >>> + >>> +// Analyze the alloca call in STMT and return an `enum alloca_type' >>> +// explaining what type of alloca it is. IS_VLA is set if the alloca >>> +// call is really a BUILT_IN_ALLOCA_WITH_ALIGN, signifying a VLA. >>> +// >>> +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is >>> +// set to the bound used to determine this. ASSUMED_LIMIT is only set >>> +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. >>> +// >>> +// If the alloca call may be too large because of a cast from a signed >>> +// type to an unsigned type, set *INVALID_CASTED_TYPE to the >>> +// problematic signed type. >>> + >>> +static enum alloca_type >>> +alloca_call_type (gimple *stmt, bool is_vla, wide_int *assumed_limit, >>> + tree *invalid_casted_type) >> Again, consider an aggregate return. > > Done. > >> >>> +{ >>> + gcc_assert (gimple_alloca_call_p (stmt)); >>> + tree len = gimple_call_arg (stmt, 0); >>> + enum alloca_type w = ALLOCA_UNBOUNDED; >>> + wide_int min, max; >>> + >>> + gcc_assert (!is_vla || warn_vla_limit > 0); >>> + gcc_assert (is_vla || warn_alloca_limit > 0); >>> + >>> + // Adjust warn_alloca_max_size for VLAs, by taking the underlying >>> + // type into account. >>> + unsigned HOST_WIDE_INT max_size; >>> + if (is_vla) >>> + max_size = (unsigned HOST_WIDE_INT) warn_vla_limit; >>> + else >>> + max_size = (unsigned HOST_WIDE_INT) warn_alloca_limit; >>> + >>> + // Check for the obviously bounded case. >>> + if (TREE_CODE (len) == INTEGER_CST) >>> + { >>> + if (tree_to_uhwi (len) > max_size) >>> + { >>> + *assumed_limit = len; >>> + return ALLOCA_BOUND_DEFINITELY_LARGE; >>> + } >>> + if (integer_zerop (len)) >>> + return ALLOCA_ARG_IS_ZERO; >>> + w = ALLOCA_OK; >>> + } >>> + else if (TREE_CODE (len) != SSA_NAME) >>> + return ALLOCA_UNBOUNDED; >> Hmm, other than INTEGER_CST and SSA_NAME, is there any other nodes we >> can get here? Perhaps we get DECLs and such, particularly when not >> optimizing?!? > > Nope. We don't even run without optimization (because we need VRP/range > info). I added a gcc_unreachable() to make sure and added an > appropriate comment. It doesn't get triggered on tests or > glibc/binutils builds. > >> >>> + // Check the range info if available. >>> + else >>> + { >>> + if (value_range_type range_type = get_range_info (len, &min, >>> &max)) >>> + { >>> + if (range_type == VR_RANGE) >>> + { >>> + if (wi::leu_p (max, max_size)) >>> + w = ALLOCA_OK; >>> + else if (is_max (len, max)) >>> + { >>> + // A cast may have created a range we don't care >>> + // about. For instance, a cast from 16-bit to >>> + // 32-bit creates a range of 0..65535, even if there >>> + // is not really a determinable range in the >>> + // underlying code. In this case, look through the >>> + // cast at the original argument, and fall through >>> + // to look at other alternatives. >>> + gimple *def = SSA_NAME_DEF_STMT (len); >>> + if (gimple_assign_cast_p (def)) >>> + len = gimple_assign_rhs1 (def); >>> + } >>> + else >>> + { >>> + /* If `len' is merely a cast that is being >>> + calculated right before the call to alloca, look >>> + at the range for the original value. >> Yea, this is similar to my comment earlier that the RHS of the cast may >> have a known range (non-negative) that allows us to not worry about the >> cast creating a huge integer. I think you can add handling that case >> here without too much trouble. Though you might consider pulling all >> the casting bits into a separate function. > > See previous comments. I've merged and cleaned this up. > >> >>> + >>> + This avoids the cast creating a range where the >>> + original expression did not have a range: >>> + >>> + # RANGE [0, 18446744073709551615] NONZERO 4294967295 >>> + _2 = (long unsigned int) n_7(D); >>> + p_9 = __builtin_alloca (_2); >> Note this example would make more sense if the type of n_7 was explicit. > > Merged and removed. > >> >> >>> + else if (range_type == VR_ANTI_RANGE) >>> + { >>> + // There may be some wrapping around going on. Catch it >>> + // with this heuristic. Hopefully, this VR_ANTI_RANGE >>> + // nonsense will go away, and we won't have to catch the >>> + // sign conversion problems with this crap. >>> + if (cast_from_signed_p (len, invalid_casted_type)) >>> + return ALLOCA_CAST_FROM_SIGNED; >> Another place where casting from an object with a non-negative range >> ought to be filtered out as not problematical. > > Same. > >> >> >>> + >>> + // If we couldn't find anything, try a few heuristics for things we >>> + // can easily determine. Check these misc cases but only accept >>> + // them if all predecessors have a known bound. >>> + basic_block bb = gimple_bb (stmt); >>> + if (w == ALLOCA_UNBOUNDED) >>> + { >>> + w = ALLOCA_OK; >>> + for (unsigned ix = 0; ix < EDGE_COUNT (bb->preds); ix++) >>> + { >>> + enum alloca_type w >>> + = alloca_call_type_by_arg (len, EDGE_PRED (bb, ix), max_size, >>> + assumed_limit); >>> + if (w != ALLOCA_OK) >>> + return w; >>> + } >> So this is an interesting tidbit that answers my questions about how we >> use alloca_call_type_by_arg. Essentially this is meant to catch the >> merge point for flow control and give a conservative warning. That's >> fine and good. But ISTM it's really a bit of a hack. What if we have >> something like this: >> >> X Y Z >> \ | / >> \|/ >> A >> / \ >> / \ >> B C >> / \ >> / \ >> D E >> >> >> Where the alloca call is in E and the incoming edges to A actually have >> useful information about the argument to the alloca call. >> >> ISTM you need to be doing something with the dominator tree here to find >> the merge point(s) where we might know something useful. And it's this >> kind of test that makes me wonder about re-purposing some of the path >> analysis code from tree-ssa-uninit.c. It may be the case that the path >> Z->A->C->E is unfeasible, but left in the CFG because duplication to >> expose the unfeasible path was unprofitable. If it turns out that the >> only argument that causes problems comes from the edge Z->A, then we >> wouldn't want to warn in this case. >> >> I don't see Andrew's work necessarily being able to solve this problem. > > In my limited testing I've seen that 95% of all cases (I'm pulling this > number out of thin air ;-)) are relatively simple. Just looking at the > definition of the SSA name in the alloca() call or the immediate > predecessors yields most of the information we need. I haven't seen > much more complicated things with an actual range. > > So I am hesitant to complicate things for something that doesn't seem as > likely to happen. Perhaps, as a follow-up if it happens in the wild? > > However, if you feel strongly about this, I will finally get around to > reading tree-ssa-uninit.c and getting down to business :). > >> >> >> >> >> >>> + } >>> + >>> + return w; >>> +} >>> + >>> +// Return TRUE if the alloca call in STMT is in a loop, otherwise >>> +// return FALSE. As an exception, ignore alloca calls for VLAs that >>> +// occur in a loop since those will be cleaned up when they go out of >>> +// scope. >>> + >>> +static bool >>> +in_loop_p (bool is_vla, gimple *stmt) >>> +{ >>> + basic_block bb = gimple_bb (stmt); >>> + if (bb->loop_father >>> + // ?? Functions with no loops get a loop_father? I >>> + // don't get it. The following conditional seems to do >>> + // the trick to exclude such nonsense. >>> + && bb->loop_father->header != ENTRY_BLOCK_PTR_FOR_FN (cfun)) >> I believe there is a "loop" that encompasses the whole function. > > Remove clueless comment. > > Phew. That took longer than expected. > > Regstrapped on x86-64 Linux and the resulting compiler was verified by > building glibc and binutils. > > Aldy
diff --git a/gcc/Makefile.in b/gcc/Makefile.in index 0786fa3..8411bed 100644 --- a/gcc/Makefile.in +++ b/gcc/Makefile.in @@ -1284,6 +1284,7 @@ OBJS = \ gimple-ssa-nonnull-compare.o \ gimple-ssa-split-paths.o \ gimple-ssa-strength-reduction.o \ + gimple-ssa-warn-alloca.o \ gimple-streamer-in.o \ gimple-streamer-out.o \ gimple-walk.o \ diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c index c11e7e7..6e82fc8 100644 --- a/gcc/c-family/c-opts.c +++ b/gcc/c-family/c-opts.c @@ -376,6 +376,16 @@ c_common_handle_option (size_t scode, const char *arg, int value, cpp_opts->warn_num_sign_change = value; break; + case OPT_Walloca_larger_than_: + if (!value) + inform (loc, "-Walloca-larger-than=0 is meaningless"); + break; + + case OPT_Wvla_larger_than_: + if (!value) + inform (loc, "-Wvla-larger-than=0 is meaningless"); + break; + case OPT_Wunknown_pragmas: /* Set to greater than 1, so that even unknown pragmas in system headers will be warned about. */ diff --git a/gcc/c-family/c.opt b/gcc/c-family/c.opt index 8c70152..017a8f9 100644 --- a/gcc/c-family/c.opt +++ b/gcc/c-family/c.opt @@ -275,6 +275,16 @@ Wall C ObjC C++ ObjC++ Warning Enable most warning messages. +Walloca +C ObjC C++ ObjC++ Var(warn_alloca) Warning +Warn on any use of alloca. + +Walloca-larger-than= +C ObjC C++ ObjC++ Var(warn_alloca_limit) Warning Joined RejectNegative UInteger +-Walloca-larger-than=<number> Warn on unbounded uses of +alloca, and on bounded uses of alloca whose bound can be larger than +<number> bytes. + Warray-bounds LangEnabledBy(C ObjC C++ ObjC++,Wall) ; in common.opt @@ -980,6 +990,12 @@ Wvla C ObjC C++ ObjC++ Var(warn_vla) Init(-1) Warning Warn if a variable length array is used. +Wvla-larger-than= +C ObjC C++ ObjC++ Var(warn_vla_limit) Warning Joined RejectNegative UInteger +-Wvla-larger-than=<number> Warn on unbounded uses of variable-length arrays, and +on bounded uses of variable-length arrays whose bound can be +larger than <number> bytes. + Wvolatile-register-var C ObjC C++ ObjC++ Var(warn_volatile_register_var) Warning LangEnabledBy(C ObjC C++ ObjC++,Wall) Warn when a register variable is declared volatile. diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi index 9a4db38..341ebb8 100644 --- a/gcc/doc/invoke.texi +++ b/gcc/doc/invoke.texi @@ -254,6 +254,7 @@ Objective-C and Objective-C++ Dialects}. @gccoptlist{-fsyntax-only -fmax-errors=@var{n} -Wpedantic @gol -pedantic-errors @gol -w -Wextra -Wall -Waddress -Waggregate-return @gol +-Walloca -Walloca-larger-than=@var{n} @gol -Wno-aggressive-loop-optimizations -Warray-bounds -Warray-bounds=@var{n} @gol -Wno-attributes -Wbool-compare -Wno-builtin-macro-redefined @gol -Wc90-c99-compat -Wc99-c11-compat @gol @@ -310,7 +311,7 @@ Objective-C and Objective-C++ Dialects}. -Wunused-const-variable -Wunused-const-variable=@var{n} @gol -Wunused-but-set-parameter -Wunused-but-set-variable @gol -Wuseless-cast -Wvariadic-macros -Wvector-operation-performance @gol --Wvla -Wvolatile-register-var -Wwrite-strings @gol +-Wvla -Wvla-larger-than=@var{n} -Wvolatile-register-var -Wwrite-strings @gol -Wzero-as-null-pointer-constant -Whsa} @item C and Objective-C-only Warning Options @@ -4660,6 +4661,61 @@ annotations. Warn about overriding virtual functions that are not marked with the override keyword. +@item -Walloca +@opindex Wno-alloca +@opindex Walloca +This option warns on all uses of @code{alloca} in the source. + +@item -Walloca-larger-than=@var{n} +This option warns on calls to @code{alloca} that are not bounded by a +controlling predicate limiting its size to @var{n} bytes, or calls to +@code{alloca} where the bound is unknown. + +For example, a bounded case of @code{alloca} could be: + +@smallexample +unsigned int n; +... +if (n <= 1000) + alloca (n); +@end smallexample + +In the above example, passing @code{-Walloca=1000} would not issue a +warning because the call to @code{alloca} is known to be at most 1000 +bytes. However, if @code{-Walloca=500} was passed, the compiler would +have emitted a warning. + +Unbounded uses, on the other hand, are uses of @code{alloca} with no +controlling predicate verifying its size. For example: + +@smallexample +stuff (); +alloca (n); +@end smallexample + +If @code{-Walloca=500} was passed, the above would trigger a warning, +but this time because of the lack of bounds checking. + +Note, that even seemingly correct code involving signed integers could +cause a warning: + +@smallexample +signed int n; +... +if (n < 500) + alloca (n); +@end smallexample + +In the above example, @var{n} could be negative, causing a larger than +expected argument to be implicitly casted into the @code{alloca} call. + +This option also warns when @code{alloca} is used in a loop. + +This warning is not enabled by @option{-Wall}, and is only active when +@option{-ftree-vrp} is active (default for @option{-O2} and above). + +See also @option{-Wvla-larger-than=@var{n}}. + @item -Warray-bounds @itemx -Warray-bounds=@var{n} @opindex Wno-array-bounds @@ -5824,9 +5880,25 @@ moving from a moved-from object, this warning can be disabled. @item -Wvla @opindex Wvla @opindex Wno-vla -Warn if variable length array is used in the code. +Warn if a variable-length array is used in the code. @option{-Wno-vla} prevents the @option{-Wpedantic} warning of -the variable length array. +the variable-length array. + +@item -Wvla-larger-than=@var{n} +If this option is used, the compiler will warn on uses of +variable-length arrays where the size is either unbounded, or bounded +by an argument that can be larger than @var{n} bytes. This is similar +to how @option{-Walloca-larger-than=@var{n}} works, but with +variable-length arrays. + +Note that GCC may optimize small variable-length arrays of a known +value into plain arrays, so this warning may not get triggered for +such arrays. + +This warning is not enabled by @option{-Wall}, and is only active when +@option{-ftree-vrp} is active (default for @option{-O2} and above). + +See also @option{-Walloca-larger-than=@var{n}}. @item -Wvolatile-register-var @opindex Wvolatile-register-var diff --git a/gcc/gimple-ssa-warn-alloca.c b/gcc/gimple-ssa-warn-alloca.c new file mode 100644 index 0000000..065a72b --- /dev/null +++ b/gcc/gimple-ssa-warn-alloca.c @@ -0,0 +1,517 @@ +/* Warn on problematic uses of alloca and variable length arrays. + Copyright (C) 2016 Free Software Foundation, Inc. + Contributed by Aldy Hernandez <aldyh@redhat.com>. + +This file is part of GCC. + +GCC is free software; you can redistribute it and/or modify it under +the terms of the GNU General Public License as published by the Free +Software Foundation; either version 3, or (at your option) any later +version. + +GCC is distributed in the hope that it will be useful, but WITHOUT ANY +WARRANTY; without even the implied warranty of MERCHANTABILITY or +FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +for more details. + +You should have received a copy of the GNU General Public License +along with GCC; see the file COPYING3. If not see +<http://www.gnu.org/licenses/>. */ + +#include "config.h" +#include "system.h" +#include "coretypes.h" +#include "backend.h" +#include "tree.h" +#include "gimple.h" +#include "tree-pass.h" +#include "ssa.h" +#include "gimple-pretty-print.h" +#include "diagnostic-core.h" +#include "fold-const.h" +#include "gimple-iterator.h" +#include "tree-ssa.h" +#include "params.h" +#include "tree-cfg.h" +#include "calls.h" +#include "cfgloop.h" + +const pass_data pass_data_walloca = { + GIMPLE_PASS, + "walloca", + OPTGROUP_NONE, + TV_NONE, + PROP_cfg, // properties_required + 0, // properties_provided + 0, // properties_destroyed + 0, // properties_start + 0, // properties_finish +}; + +class pass_walloca : public gimple_opt_pass +{ +public: + pass_walloca (gcc::context *ctxt) + : gimple_opt_pass(pass_data_walloca, ctxt), first_time_p (false) + {} + opt_pass *clone () { return new pass_walloca (m_ctxt); } + void set_pass_param (unsigned int n, bool param) + { + gcc_assert (n == 0); + first_time_p = param; + } + virtual bool gate (function *); + virtual unsigned int execute (function *); + + private: + // Set to TRUE the first time we run this pass on a function. + bool first_time_p; +}; + +bool +pass_walloca::gate (function *fun ATTRIBUTE_UNUSED) +{ + // The first time this pass is called, it is called before + // optimizations have been run and range information is unavailable, + // so we can only perform strict alloca checking. + if (first_time_p) + return warn_alloca != 0; + + return warn_alloca_limit > 0 || warn_vla_limit > 0; +} + +// Possible problematic uses of alloca. +enum alloca_type { + // Alloca argument is within known bounds that are appropriate. + ALLOCA_OK, + + // Alloca argument is KNOWN to have a value that is too large. + ALLOCA_BOUND_DEFINITELY_LARGE, + + // Alloca argument may be too large. + ALLOCA_BOUND_MAYBE_LARGE, + + // Alloca argument is bounded but of an indeterminate size. + ALLOCA_BOUND_UNKNOWN, + + // Alloca argument was casted from a signed integer. + ALLOCA_CAST_FROM_SIGNED, + + // Alloca appears in a loop. + ALLOCA_IN_LOOP, + + // Alloca argument is 0. + ALLOCA_ARG_IS_ZERO, + + // Alloca call is unbounded. That is, there is no controlling + // predicate for its argument. + ALLOCA_UNBOUNDED +}; + +// We have a few heuristics up our sleeve to determine if a call to +// alloca() is within bounds. Try them out and return the type of +// alloca call this is based on its argument. +// +// Given a known argument (ARG) to alloca() and an EDGE (E) +// calculating said argument, verify that the last statement in the BB +// in E->SRC is a gate comparing ARG to an acceptable bound for +// alloca(). See examples below. +// +// MAX_SIZE is WARN_ALLOCA= adjusted for VLAs. It is the maximum size +// in bytes we allow for arg. +// +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is +// set to the bound used to determine this. ASSUMED_LIMIT is only set +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. +// +// Returns the alloca type. + +static enum alloca_type +alloca_call_type_by_arg (tree arg, edge e, unsigned max_size, + wide_int *assumed_limit) +{ + // All the tests bellow depend on the jump being on the TRUE path. + if (!(e->flags & EDGE_TRUE_VALUE)) + return ALLOCA_UNBOUNDED; + + basic_block bb = e->src; + gimple_stmt_iterator gsi = gsi_last_bb (bb); + gimple *last = gsi_stmt (gsi); + if (!last || gimple_code (last) != GIMPLE_COND) + return ALLOCA_UNBOUNDED; + + /* Check for: + if (ARG <= N) + goto <bb 3>; + else + goto <bb 4>; + <bb 3>: + alloca(ARG); + */ + if (gimple_cond_code (last) == LE_EXPR + && gimple_cond_lhs (last) == arg) + { + if (TREE_CODE (gimple_cond_rhs (last)) == INTEGER_CST) + { + tree rhs = gimple_cond_rhs (last); + if (tree_to_uhwi (rhs) > max_size) + { + *assumed_limit = rhs; + return ALLOCA_BOUND_MAYBE_LARGE; + } + return ALLOCA_OK; + } + else + return ALLOCA_BOUND_UNKNOWN; + } + + /* Check for: + if (arg .cond. LIMIT) -or- if (LIMIT .cond. arg) + alloca(arg); + + Where LIMIT has a bound of unknown range. */ + tree limit = NULL; + if (gimple_cond_lhs (last) == arg) + limit = gimple_cond_rhs (last); + else if (gimple_cond_rhs (last) == arg) + limit = gimple_cond_lhs (last); + if (limit && TREE_CODE (limit) == SSA_NAME) + { + wide_int min, max; + value_range_type range_type = get_range_info (limit, &min, &max); + if (range_type == VR_UNDEFINED || range_type == VR_VARYING) + return ALLOCA_BOUND_UNKNOWN; + // FIXME: We could try harder here and handle a possible range + // or anti-range. Hopefully the upcoming changes to range info + // will give us finer grained info, and we can avoid somersaults + // here. + } + + return ALLOCA_UNBOUNDED; +} + +// Return TRUE if SSA's definition is a cast from a signed type. +// If so, set *INVALID_CASTED_TYPE to the signed type. + +static bool +cast_from_signed_p (tree ssa, tree *invalid_casted_type) +{ + gimple *def = SSA_NAME_DEF_STMT (ssa); + if (def + && !gimple_nop_p (def) + && gimple_assign_cast_p (def) + && !TYPE_UNSIGNED (TREE_TYPE (gimple_assign_rhs1 (def)))) + { + *invalid_casted_type = TREE_TYPE (gimple_assign_rhs1 (def)); + return true; + } + return false; +} + +// Return TURE if X has a maximum range of MAX, basically covering the +// entire domain, in which case it's no range at all. + +static bool +is_max (tree x, wide_int max) +{ + return wi::max_value (TREE_TYPE (x)) == max; +} + +// Analyze the alloca call in STMT and return an `enum alloca_type' +// explaining what type of alloca it is. IS_VLA is set if the alloca +// call is really a BUILT_IN_ALLOCA_WITH_ALIGN, signifying a VLA. +// +// If the alloca bound is determined to be too large, ASSUMED_LIMIT is +// set to the bound used to determine this. ASSUMED_LIMIT is only set +// for ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE. +// +// If the alloca call may be too large because of a cast from a signed +// type to an unsigned type, set *INVALID_CASTED_TYPE to the +// problematic signed type. + +static enum alloca_type +alloca_call_type (gimple *stmt, bool is_vla, wide_int *assumed_limit, + tree *invalid_casted_type) +{ + gcc_assert (gimple_alloca_call_p (stmt)); + tree len = gimple_call_arg (stmt, 0); + enum alloca_type w = ALLOCA_UNBOUNDED; + wide_int min, max; + + gcc_assert (!is_vla || warn_vla_limit > 0); + gcc_assert (is_vla || warn_alloca_limit > 0); + + // Adjust warn_alloca_max_size for VLAs, by taking the underlying + // type into account. + unsigned HOST_WIDE_INT max_size; + if (is_vla) + max_size = (unsigned HOST_WIDE_INT) warn_vla_limit; + else + max_size = (unsigned HOST_WIDE_INT) warn_alloca_limit; + + // Check for the obviously bounded case. + if (TREE_CODE (len) == INTEGER_CST) + { + if (tree_to_uhwi (len) > max_size) + { + *assumed_limit = len; + return ALLOCA_BOUND_DEFINITELY_LARGE; + } + if (integer_zerop (len)) + return ALLOCA_ARG_IS_ZERO; + w = ALLOCA_OK; + } + else if (TREE_CODE (len) != SSA_NAME) + return ALLOCA_UNBOUNDED; + // Check the range info if available. + else + { + if (value_range_type range_type = get_range_info (len, &min, &max)) + { + if (range_type == VR_RANGE) + { + if (wi::leu_p (max, max_size)) + w = ALLOCA_OK; + else if (is_max (len, max)) + { + // A cast may have created a range we don't care + // about. For instance, a cast from 16-bit to + // 32-bit creates a range of 0..65535, even if there + // is not really a determinable range in the + // underlying code. In this case, look through the + // cast at the original argument, and fall through + // to look at other alternatives. + gimple *def = SSA_NAME_DEF_STMT (len); + if (gimple_assign_cast_p (def)) + len = gimple_assign_rhs1 (def); + } + else + { + /* If `len' is merely a cast that is being + calculated right before the call to alloca, look + at the range for the original value. + + This avoids the cast creating a range where the + original expression did not have a range: + + # RANGE [0, 18446744073709551615] NONZERO 4294967295 + _2 = (long unsigned int) n_7(D); + p_9 = __builtin_alloca (_2); + + The correct thing would've been for the user to + use size_t, which in the case above would've been + 'long unsigned int', and everything would've + worked. But we have to catch cases where the + user is using some other compatible type for the + call argument to alloca (say unsigned short). */ + gimple *def = SSA_NAME_DEF_STMT (len); + if (gimple_assign_cast_p (def)) + { + len = gimple_assign_rhs1 (def); + range_type = get_range_info (len, &min, &max); + } + + if (range_type != VR_VARYING && is_max (len, max)) + { + // Treat a max of the entire domain as if it had no + // range info, and fall through the try other + // alternatives. + } + else + { + *assumed_limit = max; + return ALLOCA_BOUND_MAYBE_LARGE; + } + } + } + else if (range_type == VR_ANTI_RANGE) + { + // There may be some wrapping around going on. Catch it + // with this heuristic. Hopefully, this VR_ANTI_RANGE + // nonsense will go away, and we won't have to catch the + // sign conversion problems with this crap. + if (cast_from_signed_p (len, invalid_casted_type)) + return ALLOCA_CAST_FROM_SIGNED; + + // Fall thru and try other things. + } + else if (range_type == VR_VARYING) + { + // No easily determined range. Try other things. + } + } + } + + // If we couldn't find anything, try a few heuristics for things we + // can easily determine. Check these misc cases but only accept + // them if all predecessors have a known bound. + basic_block bb = gimple_bb (stmt); + if (w == ALLOCA_UNBOUNDED) + { + w = ALLOCA_OK; + for (unsigned ix = 0; ix < EDGE_COUNT (bb->preds); ix++) + { + enum alloca_type w + = alloca_call_type_by_arg (len, EDGE_PRED (bb, ix), max_size, + assumed_limit); + if (w != ALLOCA_OK) + return w; + } + } + + return w; +} + +// Return TRUE if the alloca call in STMT is in a loop, otherwise +// return FALSE. As an exception, ignore alloca calls for VLAs that +// occur in a loop since those will be cleaned up when they go out of +// scope. + +static bool +in_loop_p (bool is_vla, gimple *stmt) +{ + basic_block bb = gimple_bb (stmt); + if (bb->loop_father + // ?? Functions with no loops get a loop_father? I + // don't get it. The following conditional seems to do + // the trick to exclude such nonsense. + && bb->loop_father->header != ENTRY_BLOCK_PTR_FOR_FN (cfun)) + { + // Do not warn on VLAs occurring in a loop, since VLAs are + // guaranteed to be cleaned up when they go out of scope. + // That is, there is a corresponding __builtin_stack_restore + // at the end of the scope in which the VLA occurs. + tree fndecl = gimple_call_fn (stmt); + while (TREE_CODE (fndecl) == ADDR_EXPR) + fndecl = TREE_OPERAND (fndecl, 0); + if (DECL_BUILT_IN_CLASS (fndecl) == BUILT_IN_NORMAL + && is_vla + && DECL_FUNCTION_CODE (fndecl) == BUILT_IN_ALLOCA_WITH_ALIGN) + return false; + + return true; + } + return false; +} + +unsigned int +pass_walloca::execute (function *fun) +{ + basic_block bb; + FOR_EACH_BB_FN (bb, fun) + { + for (gimple_stmt_iterator si = gsi_start_bb (bb); !gsi_end_p (si); + gsi_next (&si)) + { + gimple *stmt = gsi_stmt (si); + location_t loc = gimple_location (stmt); + + if (!gimple_alloca_call_p (stmt)) + continue; + gcc_assert (gimple_call_num_args (stmt) >= 1); + + bool is_vla = gimple_alloca_call_p (stmt) + && gimple_call_alloca_for_var_p (as_a <gcall *> (stmt)); + + // Strict mode whining for VLAs is handled by the front-end, + // so we can safely ignore this case. Also, ignore VLAs if + // the user doesn't care about them. + if (is_vla + && (warn_vla > 0 || !warn_vla_limit)) + continue; + + if (!is_vla && (warn_alloca || !warn_alloca_limit)) + { + if (warn_alloca) + warning_at (loc, OPT_Walloca, "use of %<alloca%>"); + continue; + } + + wide_int assumed_limit + = wi::to_wide (integer_zero_node, + TYPE_PRECISION (size_type_node)); + tree invalid_casted_type = NULL; + enum alloca_type w = alloca_call_type (stmt, is_vla, &assumed_limit, + &invalid_casted_type); + + // Even if we think the alloca call is OK, make sure it's + // not in a loop. + if (w == ALLOCA_OK && in_loop_p (is_vla, stmt)) + w = ALLOCA_IN_LOOP; + + enum opt_code wcode + = is_vla ? OPT_Wvla_larger_than_ : OPT_Walloca_larger_than_; + char buff[WIDE_INT_MAX_PRECISION / 4 + 4]; + switch (w) + { + case ALLOCA_OK: + break; + case ALLOCA_BOUND_MAYBE_LARGE: + gcc_assert (assumed_limit != 0); + if (warning_at (loc, wcode, + is_vla ? "argument to variable-length array " + "may be too large" + : "argument to %<alloca%> may be too large")) + { + print_decu (assumed_limit, buff); + inform (loc, "limit is %u bytes, but argument " + "may be as large as %s", + is_vla ? warn_vla_limit : warn_alloca_limit, buff); + } + break; + case ALLOCA_BOUND_DEFINITELY_LARGE: + gcc_assert (assumed_limit != 0); + if (warning_at (loc, wcode, + is_vla ? "argument to variable-length array " + "is too large" + : "argument to %<alloca%> is too large")) + { + print_decu (assumed_limit, buff); + inform (loc, "limit is %u bytes, but argument is %s", + is_vla ? warn_vla_limit : warn_alloca_limit, buff); + } + break; + case ALLOCA_BOUND_UNKNOWN: + warning_at (loc, wcode, + is_vla ? "variable-length array bound is unknown" + : "%<alloca%> bound is unknown"); + break; + case ALLOCA_UNBOUNDED: + warning_at (loc, wcode, + is_vla ? "unbounded use of variable-length array" + : "unbounded use of %<alloca%>"); + break; + case ALLOCA_IN_LOOP: + warning_at (loc, wcode, + is_vla ? "use of variable-length array " + "within a loop" + : "use of %<alloca%> within a loop"); + break; + case ALLOCA_CAST_FROM_SIGNED: + gcc_assert (invalid_casted_type != NULL_TREE); + warning_at (loc, wcode, + is_vla ? "argument to variable-length array " + "may be too large due to " + "conversion from %qT to %qT" + : "argument to %<alloca%> may be too large due to " + "conversion from %qT to %qT", + invalid_casted_type, size_type_node); + break; + case ALLOCA_ARG_IS_ZERO: + warning_at (loc, wcode, + is_vla ? "argument to variable-length array is zero" + : "argument to %<alloca%> is zero"); + break; + default: + gcc_unreachable (); + } + } + } + return 0; +} + +gimple_opt_pass * +make_pass_walloca (gcc::context *ctxt) +{ + return new pass_walloca (ctxt); +} diff --git a/gcc/passes.def b/gcc/passes.def index 3647e90..591add2 100644 --- a/gcc/passes.def +++ b/gcc/passes.def @@ -43,6 +43,7 @@ along with GCC; see the file COPYING3. If not see NEXT_PASS (pass_warn_function_return); NEXT_PASS (pass_expand_omp); NEXT_PASS (pass_build_cgraph_edges); + NEXT_PASS (pass_walloca, /*strict_mode_p=*/true); TERMINATE_PASS_LIST (all_lowering_passes) /* Interprocedural optimization passes. */ @@ -303,6 +304,7 @@ along with GCC; see the file COPYING3. If not see NEXT_PASS (pass_simduid_cleanup); NEXT_PASS (pass_lower_vector_ssa); NEXT_PASS (pass_cse_reciprocals); + NEXT_PASS (pass_walloca, /*strict_mode_p=*/false); NEXT_PASS (pass_reassoc, false /* insert_powi_p */); NEXT_PASS (pass_strength_reduction); NEXT_PASS (pass_split_paths); diff --git a/gcc/testsuite/gcc.dg/Walloca-1.c b/gcc/testsuite/gcc.dg/Walloca-1.c new file mode 100644 index 0000000..34a20c3 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-1.c @@ -0,0 +1,63 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=2000 -O2" } */ + +#define alloca __builtin_alloca + +typedef __SIZE_TYPE__ size_t; +extern size_t strlen(const char *); + +extern void useit (char *); + +int num; + +void foo1 (size_t len, size_t len2, size_t len3) +{ + int i; + + for (i=0; i < 123; ++i) + { + char *s = alloca (566); /* { dg-warning "'alloca' within a loop" } */ + useit (s); + } + + char *s = alloca (123); + useit (s); // OK, constant argument to alloca + + s = alloca (num); // { dg-warning "large due to conversion" } + useit (s); + + s = alloca(90000); /* { dg-warning "is too large" } */ + useit (s); + + if (len < 2000) + { + s = alloca(len); // OK, bounded + useit (s); + } + + if (len + len2 < 2000) // OK, bounded + { + s = alloca(len + len2); + useit (s); + } + + if (len3 <= 2001) + { + s = alloca(len3); /* { dg-warning "may be too large" } */ + useit(s); + } +} + +void foo2 (__SIZE_TYPE__ len) +{ + // Test that a direct call to __builtin_alloca_with_align is not confused + // with a VLA. + void *p = __builtin_alloca_with_align (len, 8); // { dg-warning "unbounded use of 'alloca'" } + useit (p); +} + +void foo3 (unsigned char a) +{ + if (a == 0) + useit (__builtin_alloca (a)); // { dg-warning "argument to 'alloca' is zero" } +} diff --git a/gcc/testsuite/gcc.dg/Walloca-2.c b/gcc/testsuite/gcc.dg/Walloca-2.c new file mode 100644 index 0000000..284b34e --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-2.c @@ -0,0 +1,40 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=2000 -O2" } */ + +void f (void *); + +void +g1 (int n) +{ + void *p; + if (n > 0 && n < 2000) + p = __builtin_alloca (n); + else + p = __builtin_malloc (n); + f (p); +} + +void +g2 (int n) +{ + void *p; + if (n < 2000) + p = __builtin_alloca (n); // { dg-warning "large due to conversion" } + else + p = __builtin_malloc (n); + f (p); +} + +void +g3 (int n) +{ + void *p; + if (n > 0 && n < 3000) + { + p = __builtin_alloca (n); // { dg-warning "'alloca' may be too large" } + // { dg-message "note:.*argument may be as large as 2999" "note" { target *-*-* } 34 } + } + else + p = __builtin_malloc (n); + f (p); +} diff --git a/gcc/testsuite/gcc.dg/Walloca-3.c b/gcc/testsuite/gcc.dg/Walloca-3.c new file mode 100644 index 0000000..5345197 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-3.c @@ -0,0 +1,33 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=2000 -O2" } */ + +void f (void *); + +__SIZE_TYPE__ LIMIT; + +// Warn when there is an alloca bound, but it is an unknown bound. + +void +g1 (__SIZE_TYPE__ n) +{ + void *p; + if (n < LIMIT) + p = __builtin_alloca (n); // { dg-warning "'alloca' bound is unknown" } + else + p = __builtin_malloc (n); + f (p); +} + +// Similar to the above, but do not get confused by the upcast. + +unsigned short SHORT_LIMIT; +void +g2 (unsigned short n) +{ + void *p; + if (n < SHORT_LIMIT) + p = __builtin_alloca (n); // { dg-warning "'alloca' bound is unknown" } + else + p = __builtin_malloc (n); + f (p); +} diff --git a/gcc/testsuite/gcc.dg/Walloca-4.c b/gcc/testsuite/gcc.dg/Walloca-4.c new file mode 100644 index 0000000..d96cc4e --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-4.c @@ -0,0 +1,20 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=5000 -O2" } */ +/* { dg-xfail-if "Currently broken but Andrew's work should fix this" { *-*-* } } */ + +// Should be another variant of Walloca-7.c. +// This should not warn, as we have a known bound within limits. + + char * + _i18n_number_rewrite (char *w, char *rear_ptr) +{ + + char *src; + _Bool + use_alloca = (((rear_ptr - w) * sizeof (char)) < 4096U); + if (use_alloca) + src = (char *) __builtin_alloca ((rear_ptr - w) * sizeof (char)); + else + src = (char *) __builtin_malloc ((rear_ptr - w) * sizeof (char)); + return src; +} diff --git a/gcc/testsuite/gcc.dg/Walloca-5.c b/gcc/testsuite/gcc.dg/Walloca-5.c new file mode 100644 index 0000000..5ed1171 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-5.c @@ -0,0 +1,32 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=123 -O2" } */ +/* { dg-xfail-if "Currently broken but Andrew's work should fix this" { *-*-* } } */ + +/* The argument to alloca ends up having a range of 0..MAXINT(32bits), + so we think we have a range because of the upcast. Consequently, + we warn with "alloca may be too large", but we should technically + warn with "unbounded use of alloca". + + We currently drill through casts to figure this stuff out, but we + get confused because it's not just a cast. It's a cast, plus a + multiply. + + <bb 2>: + # RANGE [0, 4294967295] NONZERO 4294967295 + _1 = (long unsigned int) something_4(D); + # RANGE [0, 34359738360] NONZERO 34359738360 + _2 = _1 * 8; + _3 = __builtin_alloca (_2); + + I don't know whether it's even worth such fine-grained warnings. + Perhaps we should generically warn everywhere with "alloca may be + too large". + + I'm hoping that this particular case will be easier to diagnose with + Andrew's work. */ + +void useit(void *); +void foobar(unsigned int something) +{ + useit(__builtin_alloca (something * sizeof (const char *))); // { dg-warning "unbounded use of alloca" "" { xfail *-*-* } } +} diff --git a/gcc/testsuite/gcc.dg/Walloca-6.c b/gcc/testsuite/gcc.dg/Walloca-6.c new file mode 100644 index 0000000..b4d8d41 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-6.c @@ -0,0 +1,11 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=256 -O2" } */ +/* { dg-xfail-if "Currently broken but Andrew's work should fix this" { *-*-* } } */ + +void f (void*); +void g (__SIZE_TYPE__ n) +{ + // No warning on this case. Range is easily determinable. + if (n > 0 && n < 256) + f (__builtin_alloca (n)); +} diff --git a/gcc/testsuite/gcc.dg/Walloca-7.c b/gcc/testsuite/gcc.dg/Walloca-7.c new file mode 100644 index 0000000..d6581a5 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-7.c @@ -0,0 +1,17 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca -O0" } */ + +extern void f(void *); + +void foo(void) +{ + // Test that strict -Walloca works even without optimization. + f (__builtin_alloca(500)); // { dg-warning "use of 'alloca'" } +} + +void bar(void) +{ + // Test that we warn on alloca() calls, not just __builtin_alloca calls. + extern void *alloca(__SIZE_TYPE__); + f (alloca (123)); // { dg-warning "use of 'alloca'" } +} diff --git a/gcc/testsuite/gcc.dg/Walloca-8.c b/gcc/testsuite/gcc.dg/Walloca-8.c new file mode 100644 index 0000000..0d7c9c4 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Walloca-8.c @@ -0,0 +1,13 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca-larger-than=2000 -O2" } */ +/* { dg-xfail-if "Currently broken but Andrew's work should fix this" { *-*-* } } */ + +void *p; +void +foo (__SIZE_TYPE__ len) +{ + if (len < 2000 / sizeof (void *)) + p = __builtin_alloca (len * sizeof (void *)); + else + p = __builtin_malloc (len * sizeof (void *)); +} diff --git a/gcc/testsuite/gcc.dg/Wvla-1.c b/gcc/testsuite/gcc.dg/Wvla-1.c new file mode 100644 index 0000000..384c930 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Wvla-1.c @@ -0,0 +1,24 @@ +/* { dg-do compile } */ +/* { dg-options "-Wvla-larger-than=100 -O2" } */ + +typedef __SIZE_TYPE__ size_t; + +extern void useit (char *); + +int num; + +void test_vlas (size_t num) +{ + char str2[num]; /* { dg-warning "unbounded use" } */ + useit(str2); + + num = 98; + for (int i=0; i < 1234; ++i) { + char str[num]; // OK, VLA in a loop, but it is a + // known size *AND* the compiler takes + // care of cleaning up between + // iterations with + // __builtin_stack_restore. + useit(str); + } +} diff --git a/gcc/testsuite/gcc.dg/Wvla-2.c b/gcc/testsuite/gcc.dg/Wvla-2.c new file mode 100644 index 0000000..96814dc --- /dev/null +++ b/gcc/testsuite/gcc.dg/Wvla-2.c @@ -0,0 +1,70 @@ +/* { dg-do compile } */ +/* { dg-require-effective-target stdint_types } */ +/* { dg-options "-O2 -Wvla-larger-than=40" } */ + +#include <stdint.h> + +void f0 (void *); +void +f1 (__SIZE_TYPE__ a) +{ + if (a <= 10) + { + // 10 * 4 bytes = 40: OK! + uint32_t x[a]; + f0 (x); + } +} + +void +f2 (__SIZE_TYPE__ a) +{ + if (a <= 11) + { + // 11 * 4 bytes = 44: Not OK. + uint32_t x[a]; // { dg-warning "array may be too large" } + // { dg-message "note:.*argument may be as large as 44" "note" { target *-*-* } 25 } + f0 (x); + } +} + +void +f3 (__SIZE_TYPE__ a, __SIZE_TYPE__ b) +{ + if (a <= 5 && b <= 3) + { + // 5 * 3 * 4 bytes = 60: Not OK. + uint32_t x[a][b]; // { dg-warning "array may be too large" } + f0 (x); + } +} + +void +f4 (__SIZE_TYPE__ a, __SIZE_TYPE__ b) +{ + if (a <= 5 && b <= 2) + { + // 5 * 2 * 4 bytes = 40 bytes: OK! + uint32_t x[a][b]; + f0 (x); + } +} + +void +f5 (__SIZE_TYPE__ len) +{ + // Test that a direct call to __builtin_alloca_with_align is not + // confused with a VLA. + void *p = __builtin_alloca_with_align (len, 8); + f0 (p); +} + +void +f6 (unsigned stuff) +{ + int n = 7000; + do { + char a[n]; // { dg-warning "variable-length array is too large" } + f0 (a); + } while (stuff--); +} diff --git a/gcc/testsuite/gcc.dg/Wvla-3.c b/gcc/testsuite/gcc.dg/Wvla-3.c new file mode 100644 index 0000000..5124476 --- /dev/null +++ b/gcc/testsuite/gcc.dg/Wvla-3.c @@ -0,0 +1,12 @@ +/* { dg-do compile } */ +/* { dg-options "-Walloca -O2" } */ + +// Make sure we don't warn on VLA with -Walloca. + +void f (void*); + +void h1 (unsigned n) +{ + int a [n]; + f (a); +} diff --git a/gcc/tree-pass.h b/gcc/tree-pass.h index 36299a6..57b61f4 100644 --- a/gcc/tree-pass.h +++ b/gcc/tree-pass.h @@ -469,6 +469,7 @@ extern simple_ipa_opt_pass *make_pass_ipa_oacc (gcc::context *ctxt); extern simple_ipa_opt_pass *make_pass_ipa_oacc_kernels (gcc::context *ctxt); extern gimple_opt_pass *make_pass_gen_hsail (gcc::context *ctxt); extern gimple_opt_pass *make_pass_warn_nonnull_compare (gcc::context *ctxt); +extern gimple_opt_pass *make_pass_walloca (gcc::context *ctxt); /* IPA Passes */ extern simple_ipa_opt_pass *make_pass_ipa_lower_emutls (gcc::context *ctxt);