Makefile.install: fix PKGFILE expression so it works with paths with spaces

[Sid-MB]

Previously, running the make file with a PREFIX path which included spaces would write to the wrong path, now it will write to the correct path.

Also note that the second setting of PKGFILE, at line 194 previously, was useless because that line was treated as a shell command which only assigned the variable for that line's invocation.

[Copilot]

Pull request overview

This PR fixes pkg-config file generation in Makefile.install when PREFIX (and thus derived paths) contains spaces, by ensuring PKGFILE is a proper Make variable containing an unquoted path rather than a shell-quoted string.

Changes:

• Define PKGFILE as a Make variable (:=) without embedding quotes in its value, so later uses like >"$(PKGFILE)" work correctly with spaces.
• Remove the later, ineffective PKGFILE="..." assignment that was executed as a one-line shell assignment (and did not persist for subsequent recipe lines).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codspeed-hq]

Merging this PR will improve performance by 17%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

#### 🎉 Hooray! `pytest-codspeed` just leveled up to 5.0.2!

A heads-up, this is a breaking change and it might affect your current performance baseline a bit. But here's the exciting part - it's packed with new, cool features and promises improved result stability 🥳!
Curious about what's new? Visit our releases page to delve into all the awesome details about this new version.

⚡ 15 improved benchmarks
✅ 47 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	test_daxpy[100-c]	25.3 µs	21.6 µs	+17.35%
⚡	test_daxpy[100-d]	24.3 µs	20.7 µs	+17.32%
⚡	test_daxpy[100-s]	24.2 µs	20.5 µs	+18.41%
⚡	test_daxpy[100-z]	26 µs	22.2 µs	+17.26%
⚡	test_daxpy[1000-c]	32.9 µs	29.1 µs	+13.08%
⚡	test_daxpy[1000-d]	32.6 µs	28.4 µs	+14.93%
⚡	test_daxpy[1000-s]	27.7 µs	24 µs	+15.38%
⚡	test_daxpy[1000-z]	40.7 µs	37 µs	+10.11%
⚡	test_dgbmv[1-100-s]	37.4 µs	34 µs	+10.08%
⚡	test_nrm2[1000-d]	30.6 µs	27 µs	+13.57%
⚡	test_nrm2[1000-dz]	35.5 µs	31.9 µs	+11.46%
⚡	test_nrm2[100-d]	37.7 µs	25.8 µs	+46.4%
⚡	test_nrm2[100-dz]	28.9 µs	25.4 µs	+13.57%
⚡	test_dot[1000]	28.4 µs	24.5 µs	+16.12%
⚡	test_dot[100]	22.5 µs	18.1 µs	+24.17%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing Sid-MB:develop (a4aae76) with develop (3da0ff7)}