Thread: [patch] src/include/storage/s_lock.h
hi, This is a revised patch that I sent earlier to allow building pg-7.1 with gcc as well as DEC's cc. I've had good results with this applied. Could some other Alpha users try this out. Even better, could an Alpha asm guru look over the asm that I'm using (instead of the original asm in the file). in brief: the original s_lock.h file tested for __alpha and __osf__ to use the builtin __INTERLOCKED_BITSS_QUAD for our TAS. I've changed this to check for the DEC compiler (__DECC || __DECCXX), removed the __asm__(...) which appeared to be gcc-specific, and moved the whole #if (__alpha) block inside the #else of #if defined(__GNUC__). brent
Attachment
Never mind. I see the followup thread now. > hi, > > This is a revised patch that I sent earlier to allow building > pg-7.1 with gcc as well as DEC's cc. I've had good results with this > applied. Could some other Alpha users try this out. Even better, could > an Alpha asm guru look over the asm that I'm using (instead of the > original asm in the file). > > in brief: > the original s_lock.h file tested for __alpha and __osf__ to use > the builtin __INTERLOCKED_BITSS_QUAD for our TAS. I've changed this > to check for the DEC compiler (__DECC || __DECCXX), removed the > __asm__(...) which appeared to be gcc-specific, and moved the whole > #if (__alpha) block inside the #else of #if defined(__GNUC__). > > brent > [ Attachment, skipping... ] -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
Can someone comment on this? It looks OK to me. > hi, > > This is a revised patch that I sent earlier to allow building > pg-7.1 with gcc as well as DEC's cc. I've had good results with this > applied. Could some other Alpha users try this out. Even better, could > an Alpha asm guru look over the asm that I'm using (instead of the > original asm in the file). > > in brief: > the original s_lock.h file tested for __alpha and __osf__ to use > the builtin __INTERLOCKED_BITSS_QUAD for our TAS. I've changed this > to check for the DEC compiler (__DECC || __DECCXX), removed the > __asm__(...) which appeared to be gcc-specific, and moved the whole > #if (__alpha) block inside the #else of #if defined(__GNUC__). > > brent > [ Attachment, skipping... ] -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
Bruce Momjian <pgman@candle.pha.pa.us> writes: > Can someone comment on this? It looks OK to me. I've been working off-list with Brent and some of the other Alpha beta testers. I think as of a few hours ago we should pass regress tests on all Alpha platforms except NetBSD, where there seem to be some problems with NaN values. It looks like this is a NetBSD bug and not ours. The currently committed Alpha spinlock code should work if I'm reading the Alpha AXP Architecture Manual correctly, but there still seems to be concern about it in some quarters. Further testing would be a good idea, especially on multi-CPU Alphas. regards, tom lane