#
8306a37b |
| 17-Jul-2002 |
Mark Murray <markm@FreeBSD.org> |
Clean up the syntax WRT semicolons at the end of function-like-macros, and protect GCCisms from non-GNU compilers and lint.
|
Revision tags: release/4.6.0_cvs |
|
#
71acb247 |
| 28-Feb-2002 |
Bosko Milekic <bmilekic@FreeBSD.org> |
Make MPLOCKED work again in asm files and stringify it explicitly where necessary.
Reviewed by: jake
|
#
d2f22d70 |
| 11-Feb-2002 |
Bruce Evans <bde@FreeBSD.org> |
Garbage-collect the "LOCORE" version of MPLOCKED.
|
Revision tags: release/4.5.0_cvs, release/4.4.0_cvs |
|
#
e4e991e1 |
| 18-Dec-2001 |
John Baldwin <jhb@FreeBSD.org> |
Allow the ATOMIC_ASM() macro to pass in the constraints on the V parameter since the char versions need to use either ax, bx, cx, or dx.
Submitted by: Peter Jeremy (mostly) Recommended by: bde
|
#
d9009094 |
| 12-Nov-2001 |
John Baldwin <jhb@FreeBSD.org> |
Use newer constraints for atomic_cmpset().
Requested by: bde
|
#
20a2016a |
| 12-Nov-2001 |
John Baldwin <jhb@FreeBSD.org> |
Use newer constraints for inline assembly for an operand that is both an input and an output by using the '+' modifier rather than listing the operand in both the input and output sections.
Reviwed
Use newer constraints for inline assembly for an operand that is both an input and an output by using the '+' modifier rather than listing the operand in both the input and output sections.
Reviwed by: bde
show more ...
|
#
2a89a48f |
| 08-Oct-2001 |
John Baldwin <jhb@FreeBSD.org> |
Allow atomic ops to be somewhat safely used in userland. We always use lock prefixes in the userland case so that the binaries will work on both SMP and UP systems.
|
Revision tags: release/4.3.0_cvs, release/4.3.0 |
|
#
d888fc4e |
| 11-Feb-2001 |
Mark Murray <markm@FreeBSD.org> |
RIP <machine/lock.h>.
Some things needed bits of <i386/include/lock.h> - cy.c now has its own (only) copy of the COM_(UN)LOCK() macros, and IMASK_(UN)LOCK() has been moved to <i386/include/apic.h> (
RIP <machine/lock.h>.
Some things needed bits of <i386/include/lock.h> - cy.c now has its own (only) copy of the COM_(UN)LOCK() macros, and IMASK_(UN)LOCK() has been moved to <i386/include/apic.h> (AKA <machine/apic.h>). Reviewed by: jhb
show more ...
|
#
324fffae |
| 17-Jan-2001 |
John Baldwin <jhb@FreeBSD.org> |
- Sort of lie and say that %eax is an output only and not an input for the non-386 atomic_load_acq(). %eax is an input since its value is used in the cmpxchg instruction, but we don't care what
- Sort of lie and say that %eax is an output only and not an input for the non-386 atomic_load_acq(). %eax is an input since its value is used in the cmpxchg instruction, but we don't care what value it is, so setting it to a specific value is just wasteful. Thus, it is being used without being initialized as the warning stated, but it is ok for it to be used because its value isn't important. Thus, we are only sort of lying when we say it is an output only operand. - Add "cc" to the clobber list for atomic_load_acq() since the cmpxchgl changes ZF.
show more ...
|
#
8a6b1c8f |
| 16-Jan-2001 |
John Baldwin <jhb@FreeBSD.org> |
- Fix atomic_load_* and atomic_store_* to generate functions for atomic.c that modules can call. - Remove the old gcc <= 2.8 versions of the atomic ops. - Resort the order of some things in the fil
- Fix atomic_load_* and atomic_store_* to generate functions for atomic.c that modules can call. - Remove the old gcc <= 2.8 versions of the atomic ops. - Resort the order of some things in the file so that there is only one #ifdef for KLD_MODULE, and so that all WANT_FUNCTIONS stuff is moved to the bottom of the file. - Remove ATOMIC_ACQ_REL() and just use explicit macros instead.
show more ...
|
#
9d979d89 |
| 14-Jan-2001 |
John Baldwin <jhb@FreeBSD.org> |
Fix the atomic_load_acq() and atomic_store_rel() functions to properly implement memory fences for the 486+. The 386 still uses versions w/o memory fences as all operations on the 386 are not progra
Fix the atomic_load_acq() and atomic_store_rel() functions to properly implement memory fences for the 486+. The 386 still uses versions w/o memory fences as all operations on the 386 are not program ordered. The 386 versions are not MP safe.
show more ...
|
Revision tags: release/4.2.0 |
|
#
55d79ad0 |
| 28-Oct-2000 |
John Baldwin <jhb@FreeBSD.org> |
The x86 atomic operations are already locked, so they do not need an additional locked instruction to guarantee a write barrier for the acquire variants.
Approved by: dfr Pointy hat to: jhb
|
#
ee8f2f37 |
| 25-Oct-2000 |
John Baldwin <jhb@FreeBSD.org> |
- Add atomic_cmpset_{acq_,rel_,}_long - Add in atomic operations for 8-bit, 16-bit, and 32-bit integers
|
#
ccbdd9ee |
| 20-Oct-2000 |
John Baldwin <jhb@FreeBSD.org> |
- Expand the set of atomic operations to optionally include memory barriers in most of the atomic operations. Now for these operations, you can use the normal atomic operation, you can use the o
- Expand the set of atomic operations to optionally include memory barriers in most of the atomic operations. Now for these operations, you can use the normal atomic operation, you can use the operation with a read barrier, or you can use the operation with a write barrier. The function names follow the same semantics used in the ia64 instruction set. An atomic operation with a read barrier has the extra suffix 'acq', due to it having "acquire" semantics. An atomic operation with a write barrier has the extra suffix 'rel'. These suffixes are inserted between the name of the operation to perform and the typename. For example, the atomic_add_int() function now has 3 variants: - atomic_add_int() - this is the same as the previous function - atomic_add_acq_int() - this function combines the add operation with a read memory barrier - atomic_add_rel_int() - this function combines the add operation with a write memory barrier - Add 'ptr' to the list of types that we can perform atomic operations on. This allows one to do atomic operations on uintptr_t's. This is useful in the mutex code, for example, because the actual mutex lock is a pointer. - Add two new operations for doing loads and stores with memory barriers. The new load operations use a read barrier before the load, and the new store operations use a write barrier after the load. For example, atomic_load_acq_int() will atomically load an integer as well as enforcing a read barrier.
show more ...
|
#
b4645202 |
| 06-Oct-2000 |
John Baldwin <jhb@FreeBSD.org> |
Add atomic_readandclear_int and atomic_readandclear_long.
|
Revision tags: release/4.1.1_cvs |
|
#
819e370c |
| 06-Sep-2000 |
Poul-Henning Kamp <phk@FreeBSD.org> |
Introduce atomic_cmpset_int() and atomic_cmpset_long() from SMPng a few hours earlier than the rest.
The next DEVFS commit needs these functions.
Alpha versions by: dfr i386 versions by: jakeb
App
Introduce atomic_cmpset_int() and atomic_cmpset_long() from SMPng a few hours earlier than the rest.
The next DEVFS commit needs these functions.
Alpha versions by: dfr i386 versions by: jakeb
Approved by: SMPng
show more ...
|
Revision tags: release/4.1.0, release/3.5.0_cvs |
|
#
25f01030 |
| 10-May-2000 |
David E. O'Brien <obrien@FreeBSD.org> |
When using _asm{} in GCC, one must specify the operand's size if one specifies the instruction's operation size. GCC will default to 32-bit operands reguardless of the prototype (ie, formal paramete
When using _asm{} in GCC, one must specify the operand's size if one specifies the instruction's operation size. GCC will default to 32-bit operands reguardless of the prototype (ie, formal parameters' type) of an inline function.
show more ...
|
Revision tags: release/4.0.0_cvs, release/3.4.0_cvs |
|
#
693612eb |
| 04-Oct-1999 |
Peter Wemm <peter@FreeBSD.org> |
Use the rev 1.1.2.1 code from RELENG_3 for atomic operations rather than the non-atomic C macros.
|
#
b8710473 |
| 04-Oct-1999 |
Peter Wemm <peter@FreeBSD.org> |
Typo: s/__GNUC_MINOR_/__GNUC_MINOR__/ (__GNUC_MINOR__ on egcs in -current is "91" and is going to be "95" soon)
|
#
181d2137 |
| 03-Oct-1999 |
Eivind Eklund <eivind@FreeBSD.org> |
Allow compilation with older versions of GCC, in order to make it possible to bootstrap and work with -current from older versions of FreeBSD.
|
Revision tags: release/3.3.0_cvs |
|
#
c3aac50f |
| 28-Aug-1999 |
Peter Wemm <peter@FreeBSD.org> |
$Id$ -> $FreeBSD$
|
#
08c40841 |
| 18-Aug-1999 |
Alan Cox <alc@FreeBSD.org> |
Create callable (non-inline) versions of the atomic_OP_TYPE functions that are linked into the kernel. The KLD compilation options are changed to call these functions, rather than in-lining the atom
Create callable (non-inline) versions of the atomic_OP_TYPE functions that are linked into the kernel. The KLD compilation options are changed to call these functions, rather than in-lining the atomic operations.
This approach makes atomic operations from KLDs significantly faster on UP systems (though somewhat slower on SMP systems).
PR: i386/13111 Submitted by: peter.jeremy@alcatel.com.au
show more ...
|
#
03e3bc8e |
| 24-Jul-1999 |
Alan Cox <alc@FreeBSD.org> |
atomic.h: Change "void *" to "volatile TYPE *", improving type safety and eliminating some warnings (e.g., mp_machdep.c rev 1.106).
cpufunc.h: Eliminate setbits. As defined, it's not precisely c
atomic.h: Change "void *" to "volatile TYPE *", improving type safety and eliminating some warnings (e.g., mp_machdep.c rev 1.106).
cpufunc.h: Eliminate setbits. As defined, it's not precisely correct; and it's redundant. (Use atomic_set_int instead.)
ipl_funcs.c: Use atomic_set_int instead of setbits.
systm.h: Include atomic.h.
Reviewed by: bde
show more ...
|
#
47b8bc92 |
| 13-Jul-1999 |
Alan Cox <alc@FreeBSD.org> |
Commit the correct patch, i.e., the one that actually corresponds to the rev 1.2 log entry.
|
#
e58bb1c4 |
| 13-Jul-1999 |
Alan Cox <alc@FreeBSD.org> |
Changed the implementation of the primitives to guarantee atomicity with respect to interrupts on UP systems. (The upgrade from gcc 2.7.x to egcs 1.1.2 produced at least one non-atomic code sequence
Changed the implementation of the primitives to guarantee atomicity with respect to interrupts on UP systems. (The upgrade from gcc 2.7.x to egcs 1.1.2 produced at least one non-atomic code sequence in swap_pager_getpages.)
In addition, the primitives are now SMP-safe, but only on SMPs. (For portability between SMPs and UPs, modules are compiled with the SMP-safe versions.)
Submitted by: dillon and myself Reviewed by: bde
show more ...
|