| Message ID | 20210604212747.959028-1-richard.henderson@linaro.org |
|---|---|
| State | New |
| Headers | show |
| Series | tcg: Introduce tcg_remove_ops_after | expand |
On Fri, 4 Jun 2021 at 22:27, Richard Henderson <richard.henderson@linaro.org> wrote: > > Introduce a function to remove everything emitted > since a given point. > > Cc: Peter Maydell <peter.maydell@linaro.org> > Signed-off-by: Richard Henderson <richard.henderson@linaro.org> > --- > include/tcg/tcg.h | 1 + > tcg/tcg.c | 13 +++++++++++++ > 2 files changed, 14 insertions(+) > > diff --git a/include/tcg/tcg.h b/include/tcg/tcg.h > index 74cb345308..6895246fab 100644 > --- a/include/tcg/tcg.h > +++ b/include/tcg/tcg.h > @@ -1081,6 +1081,7 @@ TCGOp *tcg_emit_op(TCGOpcode opc); > void tcg_op_remove(TCGContext *s, TCGOp *op); > TCGOp *tcg_op_insert_before(TCGContext *s, TCGOp *op, TCGOpcode opc); > TCGOp *tcg_op_insert_after(TCGContext *s, TCGOp *op, TCGOpcode opc); > +void tcg_remove_ops_after(TCGOp *op); A doc comment would be nice. > void tcg_optimize(TCGContext *s); > > diff --git a/tcg/tcg.c b/tcg/tcg.c > index 0dc271aac9..262dbba1fd 100644 > --- a/tcg/tcg.c > +++ b/tcg/tcg.c > @@ -2649,6 +2649,19 @@ void tcg_op_remove(TCGContext *s, TCGOp *op) > #endif > } > > +void tcg_remove_ops_after(TCGOp *op) > +{ > + TCGContext *s = tcg_ctx; > + > + while (true) { > + TCGOp *last = tcg_last_op(); > + if (last == op) { > + return; > + } > + tcg_op_remove(s, last); > + } > +} This looks OK as far as TCG proper goes, but is it going to confuse the TCG plugin infrastructure if a frontend generates a bunch of TCG IR and then winds back the insn stream and generates something else instead ? I see some calls from tcg/ into plugin related functions, so I'm not sure... thanks -- PMM
On 6/7/21 6:51 AM, Peter Maydell wrote: >> +void tcg_remove_ops_after(TCGOp *op); > > A doc comment would be nice. Fair enough. > This looks OK as far as TCG proper goes, but is it going to confuse > the TCG plugin infrastructure if a frontend generates a bunch of > TCG IR and then winds back the insn stream and generates something > else instead ? I see some calls from tcg/ into plugin related > functions, so I'm not sure... Mm. Didn't think about that Friday. But looking now, if you wind back to dc->insn_start, everything will be fine. The plugin stuff will put stuff in the stream, and then go back and post-process it. But it does not record anything on the side, so if the stuff is no longer in the stream all's well. r~
diff --git a/include/tcg/tcg.h b/include/tcg/tcg.h index 74cb345308..6895246fab 100644 --- a/include/tcg/tcg.h +++ b/include/tcg/tcg.h @@ -1081,6 +1081,7 @@ TCGOp *tcg_emit_op(TCGOpcode opc); void tcg_op_remove(TCGContext *s, TCGOp *op); TCGOp *tcg_op_insert_before(TCGContext *s, TCGOp *op, TCGOpcode opc); TCGOp *tcg_op_insert_after(TCGContext *s, TCGOp *op, TCGOpcode opc); +void tcg_remove_ops_after(TCGOp *op); void tcg_optimize(TCGContext *s); diff --git a/tcg/tcg.c b/tcg/tcg.c index 0dc271aac9..262dbba1fd 100644 --- a/tcg/tcg.c +++ b/tcg/tcg.c @@ -2649,6 +2649,19 @@ void tcg_op_remove(TCGContext *s, TCGOp *op) #endif } +void tcg_remove_ops_after(TCGOp *op) +{ + TCGContext *s = tcg_ctx; + + while (true) { + TCGOp *last = tcg_last_op(); + if (last == op) { + return; + } + tcg_op_remove(s, last); + } +} + static TCGOp *tcg_op_alloc(TCGOpcode opc) { TCGContext *s = tcg_ctx;
Introduce a function to remove everything emitted since a given point. Cc: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/tcg/tcg.h | 1 + tcg/tcg.c | 13 +++++++++++++ 2 files changed, 14 insertions(+) -- 2.25.1