Display the source code in dmd/backend/x86/cod4.d from which thispage was generated on github.

If you spot a problem with this page, click here to create aBugzilla issue.

Quickly fork, edit online, and submit a pull request for this page.Requires a signed-in GitHub account. This works well for small changes.If you'd like to make larger changes you may want to consider usinglocal clone.

Module `dmd.backend.x86.cod4`

Code generation 4

Includes

- assignemt variations of operators (+= -= *= /= %= <<= >>=) - integer comparison (< > <= >=) - converting integers to a different size (e.g. short to int) - bit instructions (bit scan, population count)

Compiler implementation of the D programming language.

Mostly code generation for assignment operators.

Documentation

https://dlang.org/phobos/dmd_backend_x86_cod4.html

Coverage

https://codecov.io/gh/dlang/dmd/src/master/compiler/src/dmd/backend/x86/cod4.d

Functions

Name	Description
`cdaddass(cg, cdb, e, pretregs)`	Generate code for += -= &= \|= ^= negass
`cdasm(cg, cdb, e, pretregs)`	Generate code for an asm elem.
`cdbscan(cg, cdb, e, pretregs)`	Generate code for OPbsf and OPbsr.
`cdbt(cg, cdb, e, pretregs)`	Generate code for OPbt, OPbtc, OPbtr, OPbts
`cdbtst(cg, cdb, e, pretregs)`	Generate code for OPbtst
`cdbyteint(cg, cdb, e, pretregs)`	Convert byte to int. For OPu8_16 and OPs8_16.
`cdcmp(cg, cdb, e, pretregs)`	Generate code for compares. Handles lt,gt,le,ge,eqeq,ne for all data types.
`cdcmpxchg(cg, cdb, e, pretregs)`	Generate code for OPcmpxchg
`cdcnvt(cg, cdb, e, pretregs)`	Do conversions. Depends on OPd_s32 and CLIB.dbllng being in sequence. OPvp_fp OPcvp_fp OPd_s32 OPb_8 OPs32_d OPd_s16 OPs16_d OPd_u16 OPu16_d OPd_u32 OPu32_d OPd_f OPf_d OPd_ld OPld_d
`cddivass(cg, cdb, e, pretregs)`	Generate code for /= %=
`cdeq(cg, cdb, e, pretregs)`	Generate code for an assignment.
`cdfar16(cg, cdb, e, pretregs)`	Generate code for OPnp_f16p and OPf16p_np.
`cdlngsht(cg, cdb, e, pretregs)`	Convert long to short. OPoffset OP32_16 OP16_8 OP64_32 OP128_64
`cdmsw(cg, cdb, e, pretregs)`	Get top 32 bits of 64 bit value (I32) or top 16 bits of 32 bit value (I16) or top 64 bits of 128 bit value (I64). OPmsw
`cdmulass(cg, cdb, e, pretregs)`	Generate code for *=
`cdpair(cg, cdb, e, pretregs)`	Generate code for OPpair, OPrpair.
`cdpopcnt(cg, cdb, e, pretregs)`	OPpopcnt operator
`cdport(cg, cdb, e, pretregs)`	Handle operators OPinp and OPoutp.
`cdprefetch(cg, cdb, e, pretregs)`	Generate code for OPprefetch
`cdshass(cg, cdb, e, pretregs)`	Generate code for <<= and >>=
`cdshtlng(cg, cdb, e, pretregs)`	Convert short to long. OPnp_fp OPs16_32 OPu16_32 OPu32_64 OPs32_64 OPu64_128 OPs64_128
`doinreg(s, e)`	Determine if expression e can be evaluated directly into register variable s. Have to be careful about things like x=x+x+x, and x=a+x.
`longcmp(cdb, e, jcond, fltarg, targ)`	Generate code for signed compare of longs.
`modEA(cdb, c)`	Return code for saving common subexpressions if EA turns out to be a register. This is called just before modifying an EA.

Authors

Walter Bright

License

Boost License 1.0