<div dir="ltr"><span style="font-size:13px">Hi Dave,</span><div><span style="font-size:13px"><br></span></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">Not sure I follow - why do they need to end up in a zero-init section? (it seems like/I assume their value doesn't matter at all, but perhaps I'm misunderstanding)</span></blockquote><div><br></div><div>Zero initialized things can go in a zero-init section so they don't take up space on disk. Uninitialized globals are probably put in there by default though, so these initializations are probably unnecessary. I'll check, and if so remove them.</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">What I mean is that "branch to unreachable -> assert" is a simple transform someone would make, so I probably wouldn't leave some code like that that looks so easy to simplify & isn't.</span></blockquote><div><br></div><div>It's actually safe today, because the 'checked' flag is only used if NDEBUG is set, which is the same thing gating assertions. That's flimsy though - I knew I wanted a better solution long term, and I think a 'pre-checked' error with an ugly name is the right way to go.</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">Not sure of the best name... :/ I feel like something like "Error::Quiet()" or something, but that's pretty vague/subtle, probably.</span></blockquote><div><span style="font-size:13px"><br></span></div><div>Yeah - this will be pretty rare, and I want it to be ugly - nobody should use a pre-checked error except for the error-as-out-parameter idiom. Actually Error::errorAsOutParameter() sounds good. :)</div><div><br></div><div>- Lang.</div><div><span style="font-size:13px"><br></span></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Mar 25, 2016 at 11:35 AM, David Blaikie <span dir="ltr"><<a href="mailto:dblaikie@gmail.com" target="_blank">dblaikie@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote"><span class="">On Fri, Mar 25, 2016 at 11:25 AM, Lang Hames <span dir="ltr"><<a href="mailto:lhames@gmail.com" target="_blank">lhames@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hi Dave,<span><div><br></div><div>> <span style="font-size:13px">drop the empty "private:" section?</span></div><div><span style="font-size:13px"><br></span></div></span><div><span style="font-size:13px">Good call. Done in r</span>264436.</div><span><br><blockquote style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex" class="gmail_quote"><div><div style="font-size:13px"><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">+char BinaryError::ID = 0;</blockquote></div></div></div><div><div style="font-size:13px"><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">+char GenericBinaryError::ID = 0;</blockquote></div></div></div></blockquote><div> </div><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">I'm assuming these don't really need values( just their address is used) so perhaps leave them uninitialized? (rather than giving the impression their value is important)</blockquote></div><br></span>Nope. As long as they end up in a zero-init section we're fine.</div></blockquote><div><br></div></span><div>Not sure I follow - why do they need to end up in a zero-init section? (it seems like/I assume their value doesn't matter at all, but perhaps I'm misunderstanding)<br><br>Also shouldn't they be const? (well, if their value doesn't matter it shouldn't matter if their value is changed - but might want to lock it down a bit - and at that point maybe the language will force you to give it an initializer?)</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"> I usually zero-init explicitly to avoid any uninitialized warnings, but I guess those don't trigger on variables if we only take the address?</div></blockquote><div><br></div></span><div>Yeah, it'd be a pretty clear false positive if we complained about init of a variable whose value is never used (& init warnings on globals are generally hard anyway - potentially accessed from anywhere, etc)</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><span><div><br></div><blockquote style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex" class="gmail_quote"><div><div><span style="font-size:13px"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">+GenericBinaryError::GenericBinaryError(std::string FileName, Twine Msg)</blockquote></span></div></div></blockquote></span><div><span><div><span style="font-size:13px"><div><br></div></span><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Might as well pass Msg as a string instead of a twine, if there's no opportunity to delay stringification?</blockquote></div><div style="font-size:13px"><br></div></span><div>Twine also composes more freely: StringRef + {StringRef, std::string, char*} all work, and since error paths can be expensive anyway I didn't mind the overhead of creating the twine just to stringify it. I don't have strong feelings here though - switching to std::string will only mean a little bit more boilerplate at call sites.</div></div></div></blockquote><div><br></div></span><div>Yeah, mildly unfortunate that that's the tradeoff. But I can't think of any improvements to twine or string concatenation to help that greatly. It's weird that callers get infected by whether their callers are likely to have a concatenation expression or not... </div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><span><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">You can just "return {Ptr, *CmdOrErr};" if you're going to use braced init anyway/if you like.</span></blockquote><div><span style="font-size:13px"><br></span></div></span><div>The return value is an Expected<LoadCommand>, not a LoadCommand, so I have to be explicit.</div></div></blockquote><div><br></div></span><div>OK. Are we going to want a make_expected? I suppose not, since the element has to be movable in general anyway.</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">Else-after-return</span></blockquote><div><span style="font-size:13px"><br></span></div><div><span style="font-size:13px">CmdOrErr is scoped to the if - the else has to be there.</span></div><span><div><span style="font-size:13px"><br></span></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">Branch-to-unreachable should be assert. But then the error wouldn't be checked? Perhaps need a more explicit discard operation for that.</span></blockquote><div><span style="font-size:13px"><br></span></div></span><div><span style="font-size:13px">Yeah. I grappled with that for a while (I'm thinking 'assertIsSuccess'), but decided to just go with this for now and have more of a think about it.</span></div></div></blockquote><div><br></div></span><div>What I mean is that "branch to unreachable -> assert" is a simple transform someone would make, so I probably wouldn't leave some code like that that looks so easy to simplify & isn't. (our coding convention explicitly suggests unreachable over assert(false) but doesn't currently have guidance for this transformation (branch to unreachable -> assert) but I suspect it should)</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><span><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">If we're going to use Err out parameters, perhaps the caller should have some way of constructing an Err in the checked state instead of callers having to clear it?</span></blockquote><div><br></div></span><div>That's a really good idea. The 'Error as out parameter' idiom really only turns up in constructors (everything else can use Expected), so we can chose something an ugly name that people won't be tempted to abuse. How about</div><div><br></div><div>Error Err = Error::preCheckedSuccess();</div></div></blockquote><div><br></div></span><div>Not sure of the best name... :/ I feel like something like "Error::Quiet()" or something, but that's pretty vague/subtle, probably.</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><br></div><div>?</div><span><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span style="font-size:13px">Why the change to put the error checking inside the ifs here - it seems verbose/unfortunate?</span></blockquote><div><span style="font-size:13px"><br></span></div></span><div>Yep. I think the right way to fix this is to use a static named constructor to minimize the number of sites using 'error as out parameter'. I'm about to add that in a follow-up patch. </div></div></blockquote><div><br></div></span><div>Alrighty, will keep an eye out for it.</div><div><div class="h5"><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><span><font color="#888888"><div><span style="font-size:13px"><br></span></div><div><span style="font-size:13px">- Lang.</span></div><div><span style="font-size:13px"><br></span></div></font></span></div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Mar 25, 2016 at 10:53 AM, David Blaikie <span dir="ltr"><<a href="mailto:dblaikie@gmail.com" target="_blank">dblaikie@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote"><div><div>On Fri, Mar 25, 2016 at 10:25 AM, Lang Hames via llvm-commits <span dir="ltr"><<a href="mailto:llvm-commits@lists.llvm.org" target="_blank">llvm-commits@lists.llvm.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Author: lhames<br>
Date: Fri Mar 25 12:25:34 2016<br>
New Revision: 264425<br>
<br>
URL: <a href="http://llvm.org/viewvc/llvm-project?rev=264425&view=rev" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project?rev=264425&view=rev</a><br>
Log:<br>
[Object] Start threading Error through MachOObjectFile construction.<br>
<br>
Modified:<br>
llvm/trunk/include/llvm/Object/Error.h<br>
llvm/trunk/include/llvm/Object/MachO.h<br>
llvm/trunk/lib/Object/Error.cpp<br>
llvm/trunk/lib/Object/MachOObjectFile.cpp<br>
llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp<br>
<br>
Modified: llvm/trunk/include/llvm/Object/Error.h<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Object/Error.h?rev=264425&r1=264424&r2=264425&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Object/Error.h?rev=264425&r1=264424&r2=264425&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/include/llvm/Object/Error.h (original)<br>
+++ llvm/trunk/include/llvm/Object/Error.h Fri Mar 25 12:25:34 2016<br>
@@ -14,11 +14,15 @@<br>
#ifndef LLVM_OBJECT_ERROR_H<br>
#define LLVM_OBJECT_ERROR_H<br>
<br>
+#include "llvm/ADT/Twine.h"<br>
+#include "llvm/Support/Error.h"<br>
#include <system_error><br>
<br>
namespace llvm {<br>
namespace object {<br>
<br>
+class Binary;<br>
+<br>
const std::error_category &object_category();<br>
<br>
enum class object_error {<br>
@@ -39,6 +43,41 @@ inline std::error_code make_error_code(o<br>
return std::error_code(static_cast<int>(e), object_category());<br>
}<br>
<br>
+/// Base class for all errors indicating malformed binary files.<br>
+///<br>
+/// Having a subclass for all malformed binary files allows archive-walking<br>
+/// code to skip malformed files without having to understand every possible<br>
+/// way that a binary file might be malformed.<br>
+///<br>
+/// Currently inherits from ECError for easy interoperability with<br>
+/// std::error_code, but this will be removed in the future.<br>
+class BinaryError : public ErrorInfo<BinaryError, ECError> {<br></blockquote><div><br></div></div></div><div>Might as well just make this class a struct? (its publicly inheriting, has only public members) or at least drop the empty "private:" section?</div><div><div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+public:<br>
+ static char ID;<br>
+ BinaryError() {<br>
+ // Default to parse_failed, can be overridden with setErrorCode.<br>
+ setErrorCode(make_error_code(object_error::parse_failed));<br>
+ }<br>
+private:<br>
+};<br>
+<br>
+/// Generic binary error.<br>
+///<br>
+/// For errors that don't require their own specific sub-error (most errors)<br>
+/// this class can be used to describe the error via a string message.<br>
+class GenericBinaryError : public ErrorInfo<GenericBinaryError, BinaryError> {<br>
+public:<br>
+ static char ID;<br>
+ GenericBinaryError(std::string FileName, Twine Msg);<br>
+ GenericBinaryError(std::string FileName, Twine Msg, object_error ECOverride);<br>
+ const std::string &getFileName() const { return FileName; }<br>
+ const std::string &getMessage() const { return Msg; }<br>
+ void log(raw_ostream &OS) const override;<br>
+private:<br>
+ std::string FileName;<br>
+ std::string Msg;<br>
+};<br>
+<br>
} // end namespace object.<br>
<br>
} // end namespace llvm.<br>
<br>
Modified: llvm/trunk/include/llvm/Object/MachO.h<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Object/MachO.h?rev=264425&r1=264424&r2=264425&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Object/MachO.h?rev=264425&r1=264424&r2=264425&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/include/llvm/Object/MachO.h (original)<br>
+++ llvm/trunk/include/llvm/Object/MachO.h Fri Mar 25 12:25:34 2016<br>
@@ -194,7 +194,7 @@ public:<br>
typedef LoadCommandList::const_iterator load_command_iterator;<br>
<br>
MachOObjectFile(MemoryBufferRef Object, bool IsLittleEndian, bool Is64Bits,<br>
- std::error_code &EC);<br>
+ Error &Err);<br>
<br>
void moveSymbolNext(DataRefImpl &Symb) const override;<br>
<br>
<br>
Modified: llvm/trunk/lib/Object/Error.cpp<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Object/Error.cpp?rev=264425&r1=264424&r2=264425&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Object/Error.cpp?rev=264425&r1=264424&r2=264425&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/lib/Object/Error.cpp (original)<br>
+++ llvm/trunk/lib/Object/Error.cpp Fri Mar 25 12:25:34 2016<br>
@@ -58,6 +58,21 @@ std::string _object_error_category::mess<br>
"defined.");<br>
}<br>
<br>
+char BinaryError::ID = 0;<br>
+char GenericBinaryError::ID = 0;<br></blockquote><div><br></div></div></div><div>I'm assuming these don't really need values( just their address is used) so perhaps leave them uninitialized? (rather than giving the impression their value is important)</div><span><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+<br>
+GenericBinaryError::GenericBinaryError(std::string FileName, Twine Msg)<br></blockquote><div><br></div></span><div>Might as well pass Msg as a string instead of a twine, if there's no opportunity to delay stringification? (Twine is mostly used for cases where it's passed into an API that only conditionally uses it, so the concatenation can be avoided in the cases where it's not needed) This would allow moves/avoid copies, etc.</div><div><div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ : FileName(std::move(FileName)), Msg(Msg.str()) {}<br>
+<br>
+GenericBinaryError::GenericBinaryError(std::string FileName, Twine Msg, object_error ECOverride)<br>
+ : FileName(std::move(FileName)), Msg(Msg.str()) { </blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ setErrorCode(make_error_code(ECOverride));<br>
+}<br>
+<br>
+void GenericBinaryError::log(raw_ostream &OS) const {<br>
+ OS << "Error in " << FileName << ": " << Msg;<br>
+}<br>
+<br>
static ManagedStatic<_object_error_category> error_category;<br>
<br>
const std::error_category &object::object_category() {<br>
<br>
Modified: llvm/trunk/lib/Object/MachOObjectFile.cpp<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Object/MachOObjectFile.cpp?rev=264425&r1=264424&r2=264425&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Object/MachOObjectFile.cpp?rev=264425&r1=264424&r2=264425&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/lib/Object/MachOObjectFile.cpp (original)<br>
+++ llvm/trunk/lib/Object/MachOObjectFile.cpp Fri Mar 25 12:25:34 2016<br>
@@ -38,6 +38,22 @@ namespace {<br>
};<br>
}<br>
<br>
+// FIXME: Remove ECOverride once Error has been plumbed down to obj tool code.</blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+static Error<br>
+malformedError(std::string FileName, std::string Msg,<br>
+ object_error ECOverride = object_error::parse_failed) {<br>
+ return make_error<GenericBinaryError>(std::move(FileName), std::move(Msg),<br>
+ ECOverride);<br>
+}<br>
+<br>
+<br>
+// FIXME: Remove ECOverride once Error has been plumbed down to obj tool code.<br>
+static Error<br>
+malformedError(const MachOObjectFile &Obj, std::string Msg,<br>
+ object_error ECOverride = object_error::parse_failed) {<br>
+ return malformedError(Obj.getFileName(), std::move(Msg), ECOverride);<br>
+}<br>
+<br>
// FIXME: Replace all uses of this function with getStructOrErr.<br>
template <typename T><br>
static T getStruct(const MachOObjectFile *O, const char *P) {<br>
@@ -53,10 +69,10 @@ static T getStruct(const MachOObjectFile<br>
}<br>
<br>
template <typename T><br>
-static ErrorOr<T> getStructOrErr(const MachOObjectFile *O, const char *P) {<br>
+static Expected<T> getStructOrErr(const MachOObjectFile *O, const char *P) {<br>
// Don't read before the beginning or past the end of the file<br>
if (P < O->getData().begin() || P + sizeof(T) > O->getData().end())<br>
- return object_error::parse_failed;<br>
+ return malformedError(*O, "Structure read out-of-range");<br>
<br>
T Cmd;<br>
memcpy(&Cmd, P, sizeof(T));<br>
@@ -161,27 +177,25 @@ static uint32_t getSectionFlags(const Ma<br>
return Sect.flags;<br>
}<br>
<br>
-static ErrorOr<MachOObjectFile::LoadCommandInfo><br>
+static Expected<MachOObjectFile::LoadCommandInfo><br>
getLoadCommandInfo(const MachOObjectFile *Obj, const char *Ptr) {<br>
- auto CmdOrErr = getStructOrErr<MachO::load_command>(Obj, Ptr);<br>
- if (!CmdOrErr)<br>
- return CmdOrErr.getError();<br>
- if (CmdOrErr->cmdsize < 8)<br>
- return object_error::macho_small_load_command;<br>
- MachOObjectFile::LoadCommandInfo Load;<br>
- Load.Ptr = Ptr;<br>
- Load.C = CmdOrErr.get();<br>
- return Load;<br>
+ if (auto CmdOrErr = getStructOrErr<MachO::load_command>(Obj, Ptr)) {<br>
+ if (CmdOrErr->cmdsize < 8)<br>
+ return malformedError(*Obj, "Mach-O load command with size < 8 bytes",<br>
+ object_error::macho_small_load_command);<br>
+ return MachOObjectFile::LoadCommandInfo({Ptr, *CmdOrErr});<br></blockquote><div><br></div></div></div><div>You can just "return {Ptr, *CmdOrErr};" if you're going to use braced init anyway/if you like.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ } else<br></blockquote><div><br></div><div>Else-after-return</div><div><div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ return CmdOrErr.takeError();<br>
}<br>
<br>
-static ErrorOr<MachOObjectFile::LoadCommandInfo><br>
+static Expected<MachOObjectFile::LoadCommandInfo><br>
getFirstLoadCommandInfo(const MachOObjectFile *Obj) {<br>
unsigned HeaderSize = Obj->is64Bit() ? sizeof(MachO::mach_header_64)<br>
: sizeof(MachO::mach_header);<br>
return getLoadCommandInfo(Obj, getPtr(Obj, HeaderSize));<br>
}<br>
<br>
-static ErrorOr<MachOObjectFile::LoadCommandInfo><br>
+static Expected<MachOObjectFile::LoadCommandInfo><br>
getNextLoadCommandInfo(const MachOObjectFile *Obj,<br>
const MachOObjectFile::LoadCommandInfo &L) {<br>
return getLoadCommandInfo(Obj, L.Ptr + L.C.cmdsize);<br>
@@ -189,92 +203,104 @@ getNextLoadCommandInfo(const MachOObject<br>
<br>
template <typename T><br>
static void parseHeader(const MachOObjectFile *Obj, T &Header,<br>
- std::error_code &EC) {<br>
- auto HeaderOrErr = getStructOrErr<T>(Obj, getPtr(Obj, 0));<br>
- if (HeaderOrErr)<br>
- Header = HeaderOrErr.get();<br>
+ Error &Err) {<br>
+ if (auto HeaderOrErr = getStructOrErr<T>(Obj, getPtr(Obj, 0)))<br>
+ Header = *HeaderOrErr;<br>
else<br>
- EC = HeaderOrErr.getError();<br>
+ Err = HeaderOrErr.takeError();<br>
}<br>
<br>
// Parses LC_SEGMENT or LC_SEGMENT_64 load command, adds addresses of all<br>
// sections to \param Sections, and optionally sets<br>
// \param IsPageZeroSegment to true.<br>
template <typename SegmentCmd><br>
-static std::error_code parseSegmentLoadCommand(<br>
+static Error parseSegmentLoadCommand(<br>
const MachOObjectFile *Obj, const MachOObjectFile::LoadCommandInfo &Load,<br>
SmallVectorImpl<const char *> &Sections, bool &IsPageZeroSegment) {<br>
const unsigned SegmentLoadSize = sizeof(SegmentCmd);<br>
if (Load.C.cmdsize < SegmentLoadSize)<br>
- return object_error::macho_load_segment_too_small;<br>
- auto SegOrErr = getStructOrErr<SegmentCmd>(Obj, Load.Ptr);<br>
- if (!SegOrErr)<br>
- return SegOrErr.getError();<br>
- SegmentCmd S = SegOrErr.get();<br>
- const unsigned SectionSize =<br>
+ return malformedError(*Obj,<br>
+ "Mach-O segment load command size is too small",<br>
+ object_error::macho_load_segment_too_small);<br>
+ if (auto SegOrErr = getStructOrErr<SegmentCmd>(Obj, Load.Ptr)) {<br>
+ SegmentCmd S = SegOrErr.get();<br>
+ const unsigned SectionSize =<br>
Obj->is64Bit() ? sizeof(MachO::section_64) : sizeof(MachO::section);<br>
- if (S.nsects > std::numeric_limits<uint32_t>::max() / SectionSize ||<br>
- S.nsects * SectionSize > Load.C.cmdsize - SegmentLoadSize)<br>
- return object_error::macho_load_segment_too_many_sections;<br>
- for (unsigned J = 0; J < S.nsects; ++J) {<br>
- const char *Sec = getSectionPtr(Obj, Load, J);<br>
- Sections.push_back(Sec);<br>
- }<br>
- IsPageZeroSegment |= StringRef("__PAGEZERO").equals(S.segname);<br>
- return std::error_code();<br>
+ if (S.nsects > std::numeric_limits<uint32_t>::max() / SectionSize ||<br>
+ S.nsects * SectionSize > Load.C.cmdsize - SegmentLoadSize)<br>
+ return malformedError(*Obj,<br>
+ "Mach-O segment load command contains too many "<br>
+ "sections",<br>
+ object_error::macho_load_segment_too_many_sections);<br>
+ for (unsigned J = 0; J < S.nsects; ++J) {<br>
+ const char *Sec = getSectionPtr(Obj, Load, J);<br>
+ Sections.push_back(Sec);<br>
+ }<br>
+ IsPageZeroSegment |= StringRef("__PAGEZERO").equals(S.segname);<br>
+ } else<br>
+ return SegOrErr.takeError();<br>
+<br>
+ return Error::success();<br>
}<br>
<br>
MachOObjectFile::MachOObjectFile(MemoryBufferRef Object, bool IsLittleEndian,<br>
- bool Is64bits, std::error_code &EC)<br>
+ bool Is64bits, Error &Err)<br>
: ObjectFile(getMachOType(IsLittleEndian, Is64bits), Object),<br>
SymtabLoadCmd(nullptr), DysymtabLoadCmd(nullptr),<br>
DataInCodeLoadCmd(nullptr), LinkOptHintsLoadCmd(nullptr),<br>
DyldInfoLoadCmd(nullptr), UuidLoadCmd(nullptr),<br>
HasPageZeroSegment(false) {<br>
+<br>
+ // We have to check Err before it's assigned to.<br>
+ if (Err)<br>
+ llvm_unreachable("Err should be in success state at entry to constructor.");<br></blockquote><div><br></div></div></div><div>Branch-to-unreachable should be assert. But then the error wouldn't be checked? Perhaps need a more explicit discard operation for that.<br><br>If we're going to use Err out parameters, perhaps the caller should have some way of constructing an Err in the checked state instead of callers having to clear it?<br><br>(alternatively this API could be refactored into a named ctor style? (returning Expected<MachOObjectFile> and doing most of this logic in a static function rather than in the ctor? (& a private ctor that allows the processed data to be passed in to initialize the object)))</div><div><div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+<br>
if (is64Bit())<br>
- parseHeader(this, Header64, EC);<br>
+ parseHeader(this, Header64, Err);<br>
else<br>
- parseHeader(this, Header, EC);<br>
- if (EC)<br>
+ parseHeader(this, Header, Err);<br>
+ if (Err)<br>
return;<br>
<br>
uint32_t LoadCommandCount = getHeader().ncmds;<br>
if (LoadCommandCount == 0)<br>
return;<br>
<br>
- auto LoadOrErr = getFirstLoadCommandInfo(this);<br>
- if (!LoadOrErr) {<br>
- EC = LoadOrErr.getError();<br>
+ LoadCommandInfo Load;<br>
+ if (auto LoadOrErr = getFirstLoadCommandInfo(this))<br>
+ Load = *LoadOrErr;<br>
+ else {<br>
+ Err = LoadOrErr.takeError();<br>
return;<br>
}<br>
- LoadCommandInfo Load = LoadOrErr.get();<br>
+<br>
for (unsigned I = 0; I < LoadCommandCount; ++I) {<br>
LoadCommands.push_back(Load);<br>
if (Load.C.cmd == MachO::LC_SYMTAB) {<br>
// Multiple symbol tables<br>
if (SymtabLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple symbol tables");<br>
return;<br>
}<br>
SymtabLoadCmd = Load.Ptr;<br>
} else if (Load.C.cmd == MachO::LC_DYSYMTAB) {<br>
// Multiple dynamic symbol tables<br>
if (DysymtabLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple dynamic symbol tables");<br>
return;<br>
}<br>
DysymtabLoadCmd = Load.Ptr;<br>
} else if (Load.C.cmd == MachO::LC_DATA_IN_CODE) {<br>
// Multiple data in code tables<br>
if (DataInCodeLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple data-in-code tables");<br>
return;<br>
}<br>
DataInCodeLoadCmd = Load.Ptr;<br>
} else if (Load.C.cmd == MachO::LC_LINKER_OPTIMIZATION_HINT) {<br>
// Multiple linker optimization hint tables<br>
if (LinkOptHintsLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple linker optimization hint tables");<br>
return;<br>
}<br>
LinkOptHintsLoadCmd = Load.Ptr;<br>
@@ -282,24 +308,24 @@ MachOObjectFile::MachOObjectFile(MemoryB<br>
Load.C.cmd == MachO::LC_DYLD_INFO_ONLY) {<br>
// Multiple dyldinfo load commands<br>
if (DyldInfoLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple dyldinfo load commands");<br>
return;<br>
}<br>
DyldInfoLoadCmd = Load.Ptr;<br>
} else if (Load.C.cmd == MachO::LC_UUID) {<br>
// Multiple UUID load commands<br>
if (UuidLoadCmd) {<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this, "Multiple UUID load commands");<br>
return;<br>
}<br>
UuidLoadCmd = Load.Ptr;<br>
} else if (Load.C.cmd == MachO::LC_SEGMENT_64) {<br>
- if ((EC = parseSegmentLoadCommand<MachO::segment_command_64>(<br>
- this, Load, Sections, HasPageZeroSegment)))<br>
+ if ((Err = parseSegmentLoadCommand<MachO::segment_command_64>(<br>
+ this, Load, Sections, HasPageZeroSegment)))<br>
return;<br>
} else if (Load.C.cmd == MachO::LC_SEGMENT) {<br>
- if ((EC = parseSegmentLoadCommand<MachO::segment_command>(<br>
- this, Load, Sections, HasPageZeroSegment)))<br>
+ if ((Err = parseSegmentLoadCommand<MachO::segment_command>(<br>
+ this, Load, Sections, HasPageZeroSegment)))<br>
return;<br>
} else if (Load.C.cmd == MachO::LC_LOAD_DYLIB ||<br>
Load.C.cmd == MachO::LC_LOAD_WEAK_DYLIB ||<br>
@@ -309,19 +335,20 @@ MachOObjectFile::MachOObjectFile(MemoryB<br>
Libraries.push_back(Load.Ptr);<br>
}<br>
if (I < LoadCommandCount - 1) {<br>
- auto LoadOrErr = getNextLoadCommandInfo(this, Load);<br>
- if (!LoadOrErr) {<br>
- EC = LoadOrErr.getError();<br>
+ if (auto LoadOrErr = getNextLoadCommandInfo(this, Load))<br>
+ Load = *LoadOrErr;<br>
+ else {<br>
+ Err = LoadOrErr.takeError();<br>
return;<br>
}<br>
- Load = LoadOrErr.get();<br>
}<br>
}<br>
if (!SymtabLoadCmd) {<br>
if (DysymtabLoadCmd) {<br>
- // Diagnostic("truncated or malformed object (contains LC_DYSYMTAB load "<br>
- // "command without a LC_SYMTAB load command)");<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (contains "<br>
+ "LC_DYSYMTAB load command without a LC_SYMTAB load "<br>
+ "command)");<br>
return;<br>
}<br>
} else if (DysymtabLoadCmd) {</blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
@@ -330,49 +357,57 @@ MachOObjectFile::MachOObjectFile(MemoryB<br>
MachO::dysymtab_command Dysymtab =<br>
getStruct<MachO::dysymtab_command>(this, DysymtabLoadCmd);<br>
if (Dysymtab.nlocalsym != 0 && Dysymtab.ilocalsym > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (ilocalsym in LC_DYSYMTAB "<br>
- // "load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (iolocalsym in "<br>
+ "LC_DYSYMTAB load command extends past the end of "<br>
+ "the symbol table)");<br>
return;<br>
}<br>
uint64_t big_size = Dysymtab.ilocalsym;<br>
big_size += Dysymtab.nlocalsym;<br>
if (Dysymtab.nlocalsym != 0 && big_size > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (ilocalsym plus nlocalsym "<br>
- // "in LC_DYSYMTAB load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (ilocalsym plus "<br>
+ "nlocalsym in LC_DYSYMTAB load command extends past "<br>
+ "the end of the symbol table)");<br>
return;<br>
}<br>
if (Dysymtab.nextdefsym != 0 && Dysymtab.ilocalsym > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (nextdefsym in LC_DYSYMTAB "<br>
- // "load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (nextdefsym in "<br>
+ "LC_DYSYMTAB load command extends past the end of "<br>
+ "the symbol table)");<br>
return;<br>
}<br>
big_size = Dysymtab.iextdefsym;<br>
big_size += Dysymtab.nextdefsym;<br>
if (Dysymtab.nextdefsym != 0 && big_size > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (iextdefsym plus nextdefsym "<br>
- // "in LC_DYSYMTAB load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (iextdefsym plus "<br>
+ "nextdefsym in LC_DYSYMTAB load command extends "<br>
+ "past the end of the symbol table)");<br>
return;<br>
}<br>
if (Dysymtab.nundefsym != 0 && Dysymtab.iundefsym > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (nundefsym in LC_DYSYMTAB "<br>
- // "load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (nundefsym in "<br>
+ "LC_DYSYMTAB load command extends past the end of "<br>
+ "the symbol table)");<br>
return;<br>
}<br>
big_size = Dysymtab.iundefsym;<br>
big_size += Dysymtab.nundefsym;<br>
if (Dysymtab.nundefsym != 0 && big_size > Symtab.nsyms) {<br>
- // Diagnostic("truncated or malformed object (iundefsym plus nundefsym "<br>
- // "in LC_DYSYMTAB load command extends past the end of the symbol table)"<br>
- EC = object_error::parse_failed;<br>
+ Err = malformedError(*this,<br>
+ "truncated or malformed object (iundefsym plus "<br>
+ "nundefsym in LC_DYSYMTAB load command extends past "<br>
+ "the end of the symbol table");<br>
return;<br>
}<br>
}<br>
assert(LoadCommands.size() == LoadCommandCount);<br>
+<br>
+ Err = Error::success();<br>
}<br>
<br>
void MachOObjectFile::moveSymbolNext(DataRefImpl &Symb) const {<br>
@@ -2381,20 +2416,29 @@ bool MachOObjectFile::isRelocatableObjec<br>
ErrorOr<std::unique_ptr<MachOObjectFile>><br>
ObjectFile::createMachOObjectFile(MemoryBufferRef Buffer) {<br>
StringRef Magic = Buffer.getBuffer().slice(0, 4);<br>
- std::error_code EC;<br>
std::unique_ptr<MachOObjectFile> Ret;<br>
- if (Magic == "\xFE\xED\xFA\xCE")<br>
- Ret.reset(new MachOObjectFile(Buffer, false, false, EC));<br>
- else if (Magic == "\xCE\xFA\xED\xFE")<br>
- Ret.reset(new MachOObjectFile(Buffer, true, false, EC));<br>
- else if (Magic == "\xFE\xED\xFA\xCF")<br>
- Ret.reset(new MachOObjectFile(Buffer, false, true, EC));<br>
- else if (Magic == "\xCF\xFA\xED\xFE")<br>
- Ret.reset(new MachOObjectFile(Buffer, true, true, EC));<br>
- else<br>
+ if (Magic == "\xFE\xED\xFA\xCE") {<br>
+ Error Err;<br>
+ Ret.reset(new MachOObjectFile(Buffer, false, false, Err));<br></blockquote><div><br></div></div></div><div>I would generally favor MakeUnique, but I see this is existing code (& perhaps the ctor is private/protected/somesuch)<br><br>Why the change to put the error checking inside the ifs here - it seems verbose/unfortunate?<br><br>(I guess if you put it outside then you end up prematurely raising the "checked" flag on the error before it's returned? Is there some way to fix that, API-wise? have a way to check-without-clearing?)</div><div><div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ if (Err)<br>
+ return errorToErrorCode(std::move(Err));<br>
+ } else if (Magic == "\xCE\xFA\xED\xFE") {<br>
+ Error Err;<br>
+ Ret.reset(new MachOObjectFile(Buffer, true, false, Err));<br>
+ if (Err)<br>
+ return errorToErrorCode(std::move(Err));<br>
+ } else if (Magic == "\xFE\xED\xFA\xCF") {<br>
+ Error Err;<br>
+ Ret.reset(new MachOObjectFile(Buffer, false, true, Err));<br>
+ if (Err)<br>
+ return errorToErrorCode(std::move(Err));<br>
+ } else if (Magic == "\xCF\xFA\xED\xFE") {<br>
+ Error Err;<br>
+ Ret.reset(new MachOObjectFile(Buffer, true, true, Err));<br>
+ if (Err)<br>
+ return errorToErrorCode(std::move(Err));<br>
+ } else<br>
return object_error::parse_failed;<br>
<br>
- if (EC)<br>
- return EC;<br>
return std::move(Ret);<br>
}<br>
<br>
Modified: llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp?rev=264425&r1=264424&r2=264425&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp?rev=264425&r1=264424&r2=264425&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp (original)<br>
+++ llvm/trunk/tools/llvm-rtdyld/llvm-rtdyld.cpp Fri Mar 25 12:25:34 2016<br>
@@ -254,9 +254,9 @@ uint8_t *TrivialMemoryManager::allocateD<br>
<br>
static const char *ProgramName;<br>
<br>
-static int Error(const Twine &Msg) {<br>
+static int ErrorAndExit(const Twine &Msg) {<br>
errs() << ProgramName << ": error: " << Msg << "\n";<br>
- return 1;<br>
+ exit(1);<br>
}<br>
<br>
static void loadDylibs() {<br>
@@ -290,13 +290,13 @@ static int printLineInfoForInput(bool Lo<br>
ErrorOr<std::unique_ptr<MemoryBuffer>> InputBuffer =<br>
MemoryBuffer::getFileOrSTDIN(File);<br>
if (std::error_code EC = InputBuffer.getError())<br>
- return Error("unable to read input: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to read input: '" + EC.message() + "'");<br>
<br>
ErrorOr<std::unique_ptr<ObjectFile>> MaybeObj(<br>
ObjectFile::createObjectFile((*InputBuffer)->getMemBufferRef()));<br>
<br>
if (std::error_code EC = MaybeObj.getError())<br>
- return Error("unable to create object file: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to create object file: '" + EC.message() + "'");<br>
<br>
ObjectFile &Obj = **MaybeObj;<br>
<br>
@@ -309,7 +309,7 @@ static int printLineInfoForInput(bool Lo<br>
Dyld.loadObject(Obj);<br>
<br>
if (Dyld.hasError())<br>
- return Error(Dyld.getErrorString());<br>
+ ErrorAndExit(Dyld.getErrorString());<br>
<br>
// Resolve all the relocations we can.<br>
Dyld.resolveRelocations();<br>
@@ -400,19 +400,19 @@ static int executeInput() {<br>
ErrorOr<std::unique_ptr<MemoryBuffer>> InputBuffer =<br>
MemoryBuffer::getFileOrSTDIN(File);<br>
if (std::error_code EC = InputBuffer.getError())<br>
- return Error("unable to read input: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to read input: '" + EC.message() + "'");<br>
ErrorOr<std::unique_ptr<ObjectFile>> MaybeObj(<br>
ObjectFile::createObjectFile((*InputBuffer)->getMemBufferRef()));<br>
<br>
if (std::error_code EC = MaybeObj.getError())<br>
- return Error("unable to create object file: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to create object file: '" + EC.message() + "'");<br>
<br>
ObjectFile &Obj = **MaybeObj;<br>
<br>
// Load the object file<br>
Dyld.loadObject(Obj);<br>
if (Dyld.hasError()) {<br>
- return Error(Dyld.getErrorString());<br>
+ ErrorAndExit(Dyld.getErrorString());<br>
}<br>
}<br>
<br>
@@ -423,7 +423,7 @@ static int executeInput() {<br>
// Get the address of the entry point (_main by default).<br>
void *MainAddress = Dyld.getSymbolLocalAddress(EntryPoint);<br>
if (!MainAddress)<br>
- return Error("no definition for '" + EntryPoint + "'");<br>
+ ErrorAndExit("no definition for '" + EntryPoint + "'");<br>
<br>
// Invalidate the instruction cache for each loaded function.<br>
for (auto &FM : MemMgr.FunctionMemory) {<br>
@@ -432,7 +432,7 @@ static int executeInput() {<br>
// setExecutable will call InvalidateInstructionCache.<br>
std::string ErrorStr;<br>
if (!sys::Memory::setExecutable(FM, &ErrorStr))<br>
- return Error("unable to mark function executable: '" + ErrorStr + "'");<br>
+ ErrorAndExit("unable to mark function executable: '" + ErrorStr + "'");<br>
}<br>
<br>
// Dispatch to _main().<br>
@@ -452,12 +452,12 @@ static int checkAllExpressions(RuntimeDy<br>
ErrorOr<std::unique_ptr<MemoryBuffer>> CheckerFileBuf =<br>
MemoryBuffer::getFileOrSTDIN(CheckerFileName);<br>
if (std::error_code EC = CheckerFileBuf.getError())<br>
- return Error("unable to read input '" + CheckerFileName + "': " +<br>
+ ErrorAndExit("unable to read input '" + CheckerFileName + "': " +<br>
EC.message());<br>
<br>
if (!Checker.checkAllRulesInBuffer("# rtdyld-check:",<br>
CheckerFileBuf.get().get()))<br>
- return Error("some checks in '" + CheckerFileName + "' failed");<br>
+ ErrorAndExit("some checks in '" + CheckerFileName + "' failed");<br>
}<br>
return 0;<br>
}<br>
@@ -606,7 +606,7 @@ static int linkAndVerify() {<br>
<br>
// Check for missing triple.<br>
if (TripleName == "")<br>
- return Error("-triple required when running in -verify mode.");<br>
+ ErrorAndExit("-triple required when running in -verify mode.");<br>
<br>
// Look up the target and build the disassembler.<br>
Triple TheTriple(Triple::normalize(TripleName));<br>
@@ -614,29 +614,29 @@ static int linkAndVerify() {<br>
const Target *TheTarget =<br>
TargetRegistry::lookupTarget("", TheTriple, ErrorStr);<br>
if (!TheTarget)<br>
- return Error("Error accessing target '" + TripleName + "': " + ErrorStr);<br>
+ ErrorAndExit("Error accessing target '" + TripleName + "': " + ErrorStr);<br>
<br>
TripleName = TheTriple.getTriple();<br>
<br>
std::unique_ptr<MCSubtargetInfo> STI(<br>
TheTarget->createMCSubtargetInfo(TripleName, MCPU, ""));<br>
if (!STI)<br>
- return Error("Unable to create subtarget info!");<br>
+ ErrorAndExit("Unable to create subtarget info!");<br>
<br>
std::unique_ptr<MCRegisterInfo> MRI(TheTarget->createMCRegInfo(TripleName));<br>
if (!MRI)<br>
- return Error("Unable to create target register info!");<br>
+ ErrorAndExit("Unable to create target register info!");<br>
<br>
std::unique_ptr<MCAsmInfo> MAI(TheTarget->createMCAsmInfo(*MRI, TripleName));<br>
if (!MAI)<br>
- return Error("Unable to create target asm info!");<br>
+ ErrorAndExit("Unable to create target asm info!");<br>
<br>
MCContext Ctx(MAI.get(), MRI.get(), nullptr);<br>
<br>
std::unique_ptr<MCDisassembler> Disassembler(<br>
TheTarget->createMCDisassembler(*STI, Ctx));<br>
if (!Disassembler)<br>
- return Error("Unable to create disassembler!");<br>
+ ErrorAndExit("Unable to create disassembler!");<br>
<br>
std::unique_ptr<MCInstrInfo> MII(TheTarget->createMCInstrInfo());<br>
<br>
@@ -663,20 +663,20 @@ static int linkAndVerify() {<br>
MemoryBuffer::getFileOrSTDIN(Filename);<br>
<br>
if (std::error_code EC = InputBuffer.getError())<br>
- return Error("unable to read input: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to read input: '" + EC.message() + "'");<br>
<br>
ErrorOr<std::unique_ptr<ObjectFile>> MaybeObj(<br>
ObjectFile::createObjectFile((*InputBuffer)->getMemBufferRef()));<br>
<br>
if (std::error_code EC = MaybeObj.getError())<br>
- return Error("unable to create object file: '" + EC.message() + "'");<br>
+ ErrorAndExit("unable to create object file: '" + EC.message() + "'");<br>
<br>
ObjectFile &Obj = **MaybeObj;<br>
<br>
// Load the object file<br>
Dyld.loadObject(Obj);<br>
if (Dyld.hasError()) {<br>
- return Error(Dyld.getErrorString());<br>
+ ErrorAndExit(Dyld.getErrorString());<br>
}<br>
}<br>
<br>
@@ -692,7 +692,7 @@ static int linkAndVerify() {<br>
<br>
int ErrorCode = checkAllExpressions(Checker);<br>
if (Dyld.hasError())<br>
- return Error("RTDyld reported an error applying relocations:\n " +<br>
+ ErrorAndExit("RTDyld reported an error applying relocations:\n " +<br>
Dyld.getErrorString());<br>
<br>
return ErrorCode;<br>
<br>
<br>
_______________________________________________<br>
llvm-commits mailing list<br>
<a href="mailto:llvm-commits@lists.llvm.org" target="_blank">llvm-commits@lists.llvm.org</a><br>
<a href="http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits" rel="noreferrer" target="_blank">http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits</a><br>
</blockquote></div></div></div><br></div></div>
</blockquote></div><br></div>
</div></div></blockquote></div></div></div><br></div></div>
</blockquote></div><br></div>