On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
On Tue, 14 Sep 2021 15:49:36 GMT, Aleksey Shipilev wrote:
>> It isn't the "formal governance" I'm concerned about, more about the folk
>> who use/rely on Zero being the ones to evaluate the impact of a change like
>> this. People, like myself, who do not use Zero in any way cannot evaluate
>>
On Fri, 10 Sep 2021 10:02:43 GMT, David Holmes wrote:
>> Currently, the build system defaults the libjvm.so location to "server".
>> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
>> need to see if moving the libjvm.so to a proper location breaks anything.
>>
>>
On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
On Thu, 9 Sep 2021 14:31:19 GMT, Magnus Ihse Bursie wrote:
>> Currently, the build system defaults the libjvm.so location to "server".
>> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
>> need to see if moving the libjvm.so to a proper location breaks anything.
>>
On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
On Thu, 9 Sep 2021 14:31:19 GMT, Magnus Ihse Bursie wrote:
>> Currently, the build system defaults the libjvm.so location to "server".
>> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
>> need to see if moving the libjvm.so to a proper location breaks anything.
>>
On Thu, 9 Sep 2021 14:31:19 GMT, Magnus Ihse Bursie wrote:
>> Currently, the build system defaults the libjvm.so location to "server".
>> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
>> need to see if moving the libjvm.so to a proper location breaks anything.
>>
On Thu, 9 Sep 2021 15:23:45 GMT, Aleksey Shipilev wrote:
>> Sure, that works for me.
>
> While we are at it, do `core` and `custom` even carry their weight? I cannot
> remember if I ever seen anyone using them. Maybe we should "just" drop those
> variants, and leave only "server, client, minima
On Thu, 9 Sep 2021 15:12:27 GMT, Magnus Ihse Bursie wrote:
>> Ok, I agree. Can I do a Zero-specific thing here (so that it is potentially
>> cleanly backportable), and then handle the rest of the variants?
>
> Sure, that works for me.
While we are at it, do `core` and `custom` even carry their
On Thu, 9 Sep 2021 14:41:02 GMT, Aleksey Shipilev wrote:
>> I think we should stop these as well from impersonating the server JVM.
>> Preferably in the same fix, so we can remove all the special casing for
>> "server" being anything else but server.
>
> Ok, I agree. Can I do a Zero-specific th
On Thu, 9 Sep 2021 14:28:19 GMT, Magnus Ihse Bursie wrote:
>> Yes, there are at least "core" and "custom":
>>
>>
>> # All valid JVM variants
>> VALID_JVM_VARIANTS="server client minimal core zero custom"
>
> I think we should stop these as well from impersonating the server JVM.
> Preferably i
On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
On Thu, 9 Sep 2021 13:14:35 GMT, Aleksey Shipilev wrote:
>> make/autoconf/hotspot.m4 line 86:
>>
>>> 84: fi
>>> 85:
>>> 86: # All "special" variants share the same output directory ("server")
>>
>> I presume "zero" was a special variant? Are there any other variants
>> remaining?
>
> Yes,
On Thu, 9 Sep 2021 13:02:23 GMT, David Holmes wrote:
>> Currently, the build system defaults the libjvm.so location to "server".
>> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
>> need to see if moving the libjvm.so to a proper location breaks anything.
>>
>> A
On Thu, 9 Sep 2021 10:17:02 GMT, Aleksey Shipilev wrote:
> Currently, the build system defaults the libjvm.so location to "server".
> This makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We
> need to see if moving the libjvm.so to a proper location breaks anything.
>
> Ad
Currently, the build system defaults the libjvm.so location to "server". This
makes looking for `libjvm.so` awkward, see JDK-8273487 for example. We need to
see if moving the libjvm.so to a proper location breaks anything.
Additional testing:
- [x] Linux x86_64 Zero build
- [x] Linux x86_64
18 matches
Mail list logo
