Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2013-06-29 Thread Andres Freund
Hi, On 2013-06-28 23:03:22 -0400, Bruce Momjian wrote: Did we decide against specifying huge pages in Postgres? I don't think so. We need somebody to make some last modifications to the patch though and Christian doesn't seem to have the time atm. I think the bigger memory (size of the per

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2013-06-28 Thread Bruce Momjian
Did we decide against specifying huge pages in Postgres? --- On Tue, Oct 30, 2012 at 09:16:07PM +0100, Christian Kruse wrote: Hey, ok, I think I implemented all of the changes you requested. All but the ia64 dependent,

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-11-13 Thread Andres Freund
Hi CK, On 2012-10-30 21:16:07 +0100, Christian Kruse wrote: index b4fcbaf..66ed10f 100644 --- a/doc/src/sgml/config.sgml +++ b/doc/src/sgml/config.sgml I think a short introduction or at least a reference on how to configure hugepages would be a good thing. varlistentry

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-31 Thread Christian Kruse
Hey, On 30/10/12 20:33, Andres Freund wrote: +#ifdef MAP_HUGETLB +# ifdef __ia64__ +#define PG_HUGETLB_BASE_ADDR (void *)(0x8000UL) +#define PG_MAP_HUGETLB (MAP_HUGETLB|MAP_FIXED) +# else Not your fault, but that looks rather strange to me. The level of

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Christian Kruse
Hi, On 29/10/12 21:14, Peter Geoghegan wrote: I have a few initial observations on this. Thanks for your feedback. * I think you should be making the new GUC PGC_INTERNAL on platforms where MAP_HUGETLB is not defined or available. See also, effective_io_concurrency. This gives sane error

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Christian Kruse
Hey, Oh man, first I didn't sent the email to the list and now I forgot the attachment. I should really get some sleep, sorry for any inconveniences :( Greetings, CK diff --git a/doc/src/sgml/config.sgml b/doc/src/sgml/config.sgml index b4fcbaf..66ed10f 100644 --- a/doc/src/sgml/config.sgml +++

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Andres Freund
On Tuesday, October 30, 2012 08:20:33 PM Christian Kruse wrote: Hey, Oh man, first I didn't sent the email to the list and now I forgot the attachment. I should really get some sleep, sorry for any inconveniences :( +#ifdef MAP_HUGETLB +# ifdef __ia64__ +#define PG_HUGETLB_BASE_ADDR

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Christian Kruse
Hey, On 30/10/12 20:33, Andres Freund wrote: +#ifdef MAP_HUGETLB +# ifdef __ia64__ +#define PG_HUGETLB_BASE_ADDR (void *)(0x8000UL) +#define PG_MAP_HUGETLB (MAP_HUGETLB|MAP_FIXED) +# else Not your fault, but that looks rather strange to me. The level of

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Christian Kruse
Hey, ok, I think I implemented all of the changes you requested. All but the ia64 dependent, I have to do more research for this one. Greetings, CK diff --git a/doc/src/sgml/config.sgml b/doc/src/sgml/config.sgml index b4fcbaf..66ed10f 100644 --- a/doc/src/sgml/config.sgml +++

Re: [HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-30 Thread Christian Kruse
Hey, On Tuesday 30 October 2012 20:33:18 Andres Freund wrote: +#ifdef MAP_HUGETLB +# ifdef __ia64__ +#define PG_HUGETLB_BASE_ADDR (void *)(0x8000UL) +#define PG_MAP_HUGETLB (MAP_HUGETLB|MAP_FIXED) +# else Not your fault, but that looks rather strange to me. The level

[HACKERS] Re: [HACKERS] Patch für MAP_HUGETLB for mmap() shared memory

2012-10-29 Thread Peter Geoghegan
On 29 October 2012 20:14, Christian Kruse cjk+postg...@defunct.ch wrote: created a patch which implements MAP_HUGETLB for sysv shared memory segments (PGSharedMemoryCreate). It is based on tests of Tom Lane and Andres Freund, I added error handling, huge page size detection and a GUC variable.