Displaying 20 results from an estimated 5000 matches similar to: "[LLVMdev] Proposal: type uniquing of debug info for LTO"
2013 Jun 20
2
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 2:58 PM, Eric Christopher wrote:
> Hi Manman,
>
> On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>>
>> The intent of this proposal is to speedup compilation of "-flto -g" for c++ programs.
>> This is based on discussions with Adrian, David and Eric.
>>
>
> Thanks for bringing this back to the
2013 Jun 21
5
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Thu, Jun 20, 2013 at 5:25 PM, Manman Ren <mren at apple.com> wrote:
>
> On Jun 20, 2013, at 5:18 PM, David Blaikie <dblaikie at gmail.com> wrote:
>
> On Thu, Jun 20, 2013 at 5:13 PM, Manman Ren <mren at apple.com> wrote:
>
>
> On Jun 20, 2013, at 4:52 PM, David Blaikie wrote:
>
> On Thu, Jun 20, 2013 at 4:45 PM, Manman Ren <mren at apple.com>
2013 Jun 20
2
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Thu, Jun 20, 2013 at 4:45 PM, Manman Ren <mren at apple.com> wrote:
>
> On Jun 20, 2013, at 3:55 PM, Manman Ren <mren at apple.com> wrote:
>
>
> On Jun 20, 2013, at 2:58 PM, Eric Christopher wrote:
>
> Hi Manman,
>
> On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>
>
> The intent of this proposal is to speedup
2013 Jun 21
2
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Thu, Jun 20, 2013 at 5:13 PM, Manman Ren <mren at apple.com> wrote:
>
> On Jun 20, 2013, at 4:52 PM, David Blaikie wrote:
>
>> On Thu, Jun 20, 2013 at 4:45 PM, Manman Ren <mren at apple.com> wrote:
>>>
>>> On Jun 20, 2013, at 3:55 PM, Manman Ren <mren at apple.com> wrote:
>>>
>>>
>>> On Jun 20, 2013, at 2:58 PM, Eric
2013 Jun 20
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
Hi Manman,
On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>
> The intent of this proposal is to speedup compilation of "-flto -g" for c++ programs.
> This is based on discussions with Adrian, David and Eric.
>
Thanks for bringing this back to the list. The original thread was
getting quite long.
> ---------------------------
> Problem:
2013 Jun 20
1
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 3:27 PM, David Blaikie wrote:
> +1 to roughly everything Eric said.
>
> On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>>
>> The intent of this proposal is to speedup compilation of "-flto -g" for c++ programs.
>> This is based on discussions with Adrian, David and Eric.
>>
>>
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
A summary of options for issue #3:
3> To actually access the MDNode referenced via the hash value, we need to perform a lookup from the hash value to find the corresponding MDNode.
The questions are where to store this map and how to keep it up-to-date when a MDNode is replaced.
---------------------
Option a) a map in DwarfDebug, AsmWriter and DIBuilder, modify access functions in DI classes
2013 Jun 21
2
[LLVMdev] Proposal: type uniquing of debug info for LTO
Wouldn't it be simpler to name the metadata based on the hash of the
content? Then you could use a normal reference to that metadata without
needing to create a new type or teach the rest of llvm how to use it...
On Fri, Jun 21, 2013 at 3:22 PM, Manman Ren <mren at apple.com> wrote:
>
> A summary of options for issue #3:
> 3> To actually access the MDNode referenced via
2013 Jun 20
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 3:55 PM, Manman Ren <mren at apple.com> wrote:
>
> On Jun 20, 2013, at 2:58 PM, Eric Christopher wrote:
>
>> Hi Manman,
>>
>> On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>>>
>>> The intent of this proposal is to speedup compilation of "-flto -g" for c++ programs.
>>> This
2013 Jun 21
3
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Fri, Jun 21, 2013 at 11:50 AM, Manman Ren <mren at apple.com> wrote:
>
> On Jun 21, 2013, at 11:35 AM, Eric Christopher wrote:
>
>> On Thu, Jun 20, 2013 at 10:52 PM, Manman Ren <mren at apple.com> wrote:
>>>
>>> A summary of options for issue #3:
>>> 3> To actually access the MDNode referenced via the hash value, we need to perform a
2013 Jun 20
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
+1 to roughly everything Eric said.
On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren <mren at apple.com> wrote:
>
> The intent of this proposal is to speedup compilation of "-flto -g" for c++ programs.
> This is based on discussions with Adrian, David and Eric.
>
> ---------------------------
> Problem:
> A single class can be used in multiple source files and the
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 4:52 PM, David Blaikie wrote:
> On Thu, Jun 20, 2013 at 4:45 PM, Manman Ren <mren at apple.com> wrote:
>>
>> On Jun 20, 2013, at 3:55 PM, Manman Ren <mren at apple.com> wrote:
>>
>>
>> On Jun 20, 2013, at 2:58 PM, Eric Christopher wrote:
>>
>> Hi Manman,
>>
>> On Thu, Jun 20, 2013 at 2:51 PM, Manman Ren
2013 Jun 21
3
[LLVMdev] Proposal: type uniquing of debug info for LTO
More details please :]
What do you mean by "name the metadata"? Are you referring to the name field of the MDNode?
Thanks,
Manman
On Jun 21, 2013, at 7:19 AM, David Blaikie wrote:
>
> On Jun 21, 2013 1:19 AM, "Jeremy Lakeman" <Jeremy.Lakeman at gmail.com> wrote:
> >
> > Wouldn't it be simpler to name the metadata based on the hash of the content?
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 5:18 PM, David Blaikie <dblaikie at gmail.com> wrote:
> On Thu, Jun 20, 2013 at 5:13 PM, Manman Ren <mren at apple.com> wrote:
>>
>> On Jun 20, 2013, at 4:52 PM, David Blaikie wrote:
>>
>>> On Thu, Jun 20, 2013 at 4:45 PM, Manman Ren <mren at apple.com> wrote:
>>>>
>>>> On Jun 20, 2013, at 3:55 PM,
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 21, 2013, at 11:56 AM, Eric Christopher wrote:
> On Fri, Jun 21, 2013 at 11:50 AM, Manman Ren <mren at apple.com> wrote:
>>
>> On Jun 21, 2013, at 11:35 AM, Eric Christopher wrote:
>>
>>> On Thu, Jun 20, 2013 at 10:52 PM, Manman Ren <mren at apple.com> wrote:
>>>>
>>>> A summary of options for issue #3:
>>>>
2013 Jun 21
5
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Thu, Jun 20, 2013 at 10:52 PM, Manman Ren <mren at apple.com> wrote:
>
> A summary of options for issue #3:
> 3> To actually access the MDNode referenced via the hash value, we need to perform a lookup from the hash value to find the corresponding MDNode.
> The questions are where to store this map and how to keep it up-to-date when a MDNode is replaced.
>
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 20, 2013, at 5:39 PM, David Blaikie wrote:
> On Thu, Jun 20, 2013 at 5:25 PM, Manman Ren <mren at apple.com> wrote:
>>
>> On Jun 20, 2013, at 5:18 PM, David Blaikie <dblaikie at gmail.com> wrote:
>>
>> On Thu, Jun 20, 2013 at 5:13 PM, Manman Ren <mren at apple.com> wrote:
>>
>>
>> On Jun 20, 2013, at 4:52 PM, David Blaikie
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 21, 2013 1:19 AM, "Jeremy Lakeman" <Jeremy.Lakeman at gmail.com> wrote:
>
> Wouldn't it be simpler to name the metadata based on the hash of the
content? Then you could use a normal reference to that metadata without
needing to create a new type or teach the rest of llvm how to use it...
Sounds reasonable (probably with some nice prefix to get these things a
2013 Jun 21
0
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Fri, Jun 21, 2013 at 10:13 AM, Manman Ren <mren at apple.com> wrote:
>
> More details please :]
> What do you mean by "name the metadata"? Are you referring to the name field
> of the MDNode?
Using named metadata rather than unnamed metadata.
rather than having:
!llvm.hardref = !{metadata !"foo.h::myClass", !3, metadata
!"bar.h:myEnum", !4} ...
2013 Jun 21
1
[LLVMdev] Proposal: type uniquing of debug info for LTO
On Jun 21, 2013, at 10:18 AM, David Blaikie wrote:
> On Fri, Jun 21, 2013 at 10:13 AM, Manman Ren <mren at apple.com> wrote:
>>
>> More details please :]
>> What do you mean by "name the metadata"? Are you referring to the name field
>> of the MDNode?
>
> Using named metadata rather than unnamed metadata.
>
> rather than having:
>
>