<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none"><!-- p { margin-top: 0px; margin-bottom: 0px; }--></style>
</head>
<body dir="ltr" style="font-size:12pt;color:#000000;background-color:#FFFFFF;font-family:Calibri,Arial,Helvetica,sans-serif;">
<p><span style="color: rgb(33, 33, 33); font-size: 12pt;">>></span></p>
<div style="color: rgb(33, 33, 33);">
<div>
<p dir="ltr">>> Hi,<br>
>><br>
>> I was very curious so I went ahead and resubmitted Pete’s original change. So far no buildbot failures, looks promising.<br>
>><br>
>>> On 26 January 2016 at 09:19, Pete Cooper via llvm-dev<br>
>>> <<a href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a>> wrote:<br>
>>>><br>
>>>><br>
>>>> Sent from my iPhone<br>
>>>><br>
>>>>> On Jan 26, 2016, at 7:40 AM, Dave Bozier via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a>> wrote:<br>
>>>>> It would be great if there was a single<br>
>>>>><br>
>>>>> parser used by all llvm tools.<br>
>>>><br>
>>>> Absolutely. We'll get to that stage, I hope.<br>
>>>><br>
>>>> We'll need to work out if there's a way to abstract what we need from each client without slowing down or complicating any of them too much.<br>
>>><br>
>>><br>
>>> +1.<br>
>>><br>
>>> A common parser is definitely a good thing.<br>
>><br>
>><br>
>> Agree. However currently DWARFDebugFrame doesn’t have any public interface. Would it be a good first step to extract some reasonable set of accessors and back them up by unit tests?</p>
<p dir="ltr">>Maybe. For splitting the only part we need is finding where each CIE and fde is.</p>
<p dir="ltr">>George wrote the code in lld that parses bits of .eh_frame and creates .eh_frame_hdr. I think all it needs to parse is the location of the program pointer for a fde.</p>
<p dir="ltr">>So we have fairly small needs, so it would be nice to have an api that doesn't parse everything upfront.<br>
</p>
<p dir="ltr">><br>
</p>
<p dir="ltr">>Cheers,<br>
>Rafael<br>
</p>
<p dir="ltr"><br>
</p>
<p dir="ltr">For creating .eh_frame_hdr I was need to parse CIE to find <span style="background-color: rgb(255, 255, 255);">pointer encoding for the address pointers used in the FDE</span>. And then used it to read the initial PC, what was the main aim. <span style="font-size: 12pt;">E</span><span style="font-size: 12pt;">ven </span><span style="font-size: 12pt;">the
CIE </span><span style="font-size: 12pt;">augmentation string</span><span style="font-size: 12pt;"> is not fully parsed </span><span style="font-size: 12pt;">because once we found the encoding we have nothing to do there and just exit. All
</span><span style="font-size: 12pt;">unnecessary </span><span style="font-size: 12pt;">data</span><span style="font-size: 12pt;"> was just skiped during parse process. </span><span style="font-size: 12pt;">So agree with Rafael about api.</span></p>
<p dir="ltr"><br>
</p>
<p dir="ltr">George.<br>
</p>
</div>
</div>
</body>
</html>